AI Fish's Seven-Second News on Uncle Xiang's Notebook

Major AI Events

Mon, 13 Jul 2026 00:00:00 +0800

This article starts with a month-by-month record rather than a model leaderboard. Since November 2022, the AI industry has changed its questions almost every month: first, whether ordinary people could use large models; then, whether models could see, listen, generate, and reason; followed by whether they could invoke tools and operate computers; and finally, whether these capabilities could retain users, enter budgets, and generate cash flow.

The timeline only includes high-signal events that have already occurred. Model releases, open-weight releases, publicly available products, policy effective dates, and corporate governance changes are eligible; announcements, rumors, and incomplete funding rounds are not written as established facts. The first entry of each month is fixed as that month’s stock market, with a maximum of 4 entries; the remaining entries record AI industry events, through July 13, 2026 as of July.

The stock records here use Yahoo Finance’s monthly data; U.S. stocks, A-shares, and Hong Kong stocks are not forced into the same index. Each monthly table retains the previously listed high-volatility names and adds new names worth recording this month; “妖股” is a colloquial term for stocks with particularly notable monthly gains, volatility, turnover, or thematic spillover. It is neither a fundamental rating nor a buy/sell recommendation. The “monthly change” in the table is calculated based on the month-end adjusted closing price relative to the previous month’s adjusted closing price, while the “cumulative since listing” is compounded from these monthly changes and does not represent actual portfolio returns. The current month is up to July 13, 2026; for precise prices, please refer to the Yahoo historical quotes link at the end of the article.

2022: The Gateway Opens

November: ChatGPT Goes Public (NVDA +25.4%)

Market Notes: The Nasdaq continues to fluctuate amid interest rate hikes and recession expectations, while NVIDIA has rebounded from its lows, but AI has yet to form an independent market main theme; meme stocks: no clear AI meme stocks have emerged yet.

Monster Stocks Cumulative Table (As of 2022-11, Month-End Adjusted Closing Composite; NVIDIA as Leading Reference)

Target (Code)	Type	Inclusion Month	Monthly Change	Cumulative Since Inclusion
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	+25.4%	+25.4%

On November 30, OpenAI released the ChatGPT research preview. The chat interface became the first mass-market product entry point for large models, transforming model capabilities from demonstrations at research institutions into software that ordinary people could use repeatedly. OpenAI: Introducing ChatGPT

Judgment of the month: What matters is not whether ChatGPT is the first to generate text, but that natural language has for the first time become a software interface with a sufficiently low barrier to entry.

December: From Demonstration to Diffusion (NVDA -13.6%)

Market Notes: U.S. tech stocks pulled back amid rate hike expectations, with Nvidia weakening along with the semiconductor sector; the spread of ChatGPT has not yet immediately translated into stock market action, and no clear “meme stock” has emerged.

Monster Stocks Cumulative Table (as of 2022-12, month-end adjusted closing composite; Nvidia as benchmark leader)

Asset (Code)	Type	Month Added	Monthly Change	Cumulative Since Added
NVIDIA (NVDA)	Core Leader Reference	2022-11	-13.6%	+8.3%

ChatGPT enters a rapid diffusion phase. Subsequent OpenAI release notes show that users began requesting product features around context, web access, plugins, and code interpreter. The chat product was no longer just a one-off Q&A demo. ChatGPT Release Notes

Judgment for this month: The first industry increment comes from usage habits, not from new model names. Users first learn to delegate tasks to models, and only then do developers begin to look for workflows in which models can be embedded.

2023: Capabilities, Open Weights, and Capital Expenditures Accelerate Simultaneously

January: Cloud and Office Begin to Bind (BBAI +385.2%)

Market Records: US growth stocks rebounded from their 2022 lows, with Microsoft, NVIDIA, and AMD initially traded as cloud and chip recovery plays; C3.ai (AI) began to show thematic volatility, but a stable group of AI speculative stocks had not yet emerged.

Monster Stocks Cumulative Table (As of 2023-01, month-end adjusted close composite; NVIDIA as the bellwether benchmark)

Underlying (Code)	Type	Inclusion Month	Monthly Change	Cumulative Since Inclusion
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	+33.7%	+44.9%
C3.ai (AI)	U.S. High Volatility	2023-01	+77.4%	+77.4%
BigBear.ai (BBAI)	U.S. High Volatility	2023-01	+385.2%	+385.2%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	+35.2%	+35.2%

Microsoft announced the expansion of its partnership with OpenAI. The partnership places model capabilities, Azure computing power, and office software on the same commercial chain, with AI moving from a standalone product into cloud and productivity software. Microsoft: Extending the partnership with OpenAI

Verdict of the month: The advantage of closed-source models has expanded from parameter scale to cloud, distribution, enterprise contracts, and continuous services.

February: Bard and LLaMA Fork (688256.SS +121.4%)

Market Records: The competition between ChatGPT and Bard caused the AI narrative to be mapped onto stocks for the first time, with Nasdaq tech stocks strengthening; small-cap concept stocks such as C3.ai and BigBear.ai became the earliest batch of “meme stocks,” with their rises and falls driven more by sentiment and trading volume.

Monster Stock Cumulative Table (as of end of February 2023, month-end adjusted closing composite; NVIDIA as the benchmark leader)

Target (Code)	Type	Inception Month	Monthly Change	Cumulative Since Inception
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	+18.8%	+72.1%
C3.ai (AI)	US High Volatility	2023-01	+13.8%	+101.9%
BigBear.ai (BBAI)	US High Volatility	2023-01	-10.1%	+336.2%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	+110.7%	+184.9%
SoundHound AI (SOUN)	US High Volatility	2023-02	+50.3%	+50.3%
Eoptolink (300502.SZ)	A-Share High Volatility	2023-02	+50.7%	+50.7%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	+121.4%	+121.4%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	+79.4%	+79.4%

Google publicly introduces Bard. The search company begins to directly respond to ChatGPT, and generative AI enters competition for search entry points. Google: Bard and Search Updates
Meta releases the LLaMA research model. A smaller model, open research, and community reproduction provide an important starting point for the open-model ecosystem. Meta: Introducing LLaMA

Judgment of the Month: The same competition splits into two tracks: one vying for stable cloud access, and the other competing for developers, weights, and local deployment rights.

March: GPT-4 Begins Connecting to Tools (300308.SZ +52.8%)

Market Records: Under the impact of banking risks, capital instead concentrated on large-cap tech stocks and computing power, with NVIDIA becoming the main theme around GTC; A-share stocks such as 360, iFlytek, Cambricon, and Inspur showed significant ChatGPT-concept volatility.

Monstrous Stocks Cumulative Table (as of 2023-03, month-end adjusted closing price compounding; NVIDIA included as the bellwether benchmark)

Target (Code)	Type	Listed Month	Monthly Change	Cumulative Since Listing
NVIDIA (NVDA)	Core Bellwether	2022-11	+19.7%	+106.0%
C3.ai (AI)	US High-Volatility	2023-01	+48.7%	+200.2%
BigBear.ai (BBAI)	US High-Volatility	2023-01	-17.0%	+262.0%
Kunlun Tech (300418.SZ)	A-Share High-Volatility	2023-01	+39.3%	+296.8%
SoundHound AI (SOUN)	US High-Volatility	2023-02	-7.7%	+38.7%
Eoptolink (300502.SZ)	A-Share High-Volatility	2023-02	+42.2%	+114.3%
Cambricon (688256.SS)	A-Share High-Volatility	2023-02	+34.2%	+197.1%
Foxconn Industrial Internet (601138.SS)	A-Share High-Volatility	2023-02	-6.4%	+67.9%
Innolight (300308.SZ)	A-Share High-Volatility	2023-03	+52.8%	+52.8%

March 14, GPT-4 launched for ChatGPT Plus. The discussion shifted from “can it write” to complex instructions, code, reasoning, and professional tasks. OpenAI: GPT-4 research
March 23, ChatGPT began offering plugin experiments. Browsing, code interpreter, and third-party plugins pushed the model from responder to tool invoker. ChatGPT release notes

Judgment of the month: The value of a model’s capabilities increasingly depends on access to up-to-date information, computing environments, and external services.

April: Google DeepMind Merger (300476.SZ +52.8%)

Market Record: Leading US AI stocks are oscillating at high levels, while A-shares are shifting from broad-based gains to thematic rotation, with concept stocks such as iFlytek, Kunlun Tech, and 360 experiencing large volatility; “demon stocks” are starting to expand from model companies to the computing power and application chains.

Monster Stocks Cumulative Table (as of 2023-04, month-end adjusted close composite; NVIDIA as the leading benchmark)

Target (Code)	Type	Month Listed	Monthly Change	Cumulative Since Listing
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	-0.1%	+105.8%
C3.ai (AI)	US Stock – High Volatility	2023-01	-46.9%	+59.4%
BigBear.ai (BBAI)	US Stock – High Volatility	2023-01	+18.9%	+330.5%
Kunlun Tech (300418.SZ)	A-Share – High Volatility	2023-01	-12.0%	+249.2%
SoundHound AI (SOUN)	US Stock – High Volatility	2023-02	-3.6%	+33.7%
Eoptolink Technology (300502.SZ)	A-Share – High Volatility	2023-02	+22.4%	+162.3%
Cambricon (688256.SS)	A-Share – High Volatility	2023-02	+1.9%	+202.8%
Foxconn Industrial Internet (601138.SS)	A-Share – High Volatility	2023-02	+12.0%	+88.1%
Innolight (300308.SZ)	A-Share – High Volatility	2023-03	+19.4%	+82.4%
Victory Giant Technology (300476.SZ)	A-Share – High Volatility	2023-04	+52.8%	+52.8%

Google merges Google Brain with DeepMind to form Google DeepMind. Research teams, compute resources, and product roadmaps are now housed under a single organization, with Google clearly positioning the next-generation multimodal model as its core focus. Google: Formation of Google DeepMind

Judgment of the Month: Model competition is not only about research breakthroughs; it also requires organizing talent, computing power, and product distribution.

May: Long Context Meets GPU Orders (SMCI +112.4%)

Market Record: NVIDIA’s earnings report turned AI orders into a price signal for the entire market. On May 25, the stock price surged approximately 24% in a single day, and NVIDIA, AMD, Broadcom, as well as the A-share optical module chain were all追捧 together; this was the first batch of AI market movements with truly global diffusion effects.

Monster Stocks Cumulative Table (As of 2023-05, Month-end Adjusted Closing Price Compound; NVIDIA as Benchmark Leader)

Target (Code)	Type	Inclusion Month	Monthly Change	Cumulative Since Inclusion
NVIDIA (NVDA)	Core Bellwether	2022-11	+36.3%	+180.5%
C3.ai (AI)	US High Volatility	2023-01	+124.5%	+257.9%
BigBear.ai (BBAI)	US High Volatility	2023-01	-23.8%	+228.0%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	-29.8%	+145.1%
SoundHound AI (SOUN)	US High Volatility	2023-02	+14.7%	+53.4%
Eoptolink (300502.SZ)	A-Share High Volatility	2023-02	+13.0%	+196.4%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	-26.0%	+124.0%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	+39.6%	+162.5%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	+37.3%	+150.5%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	-13.9%	+31.6%
Super Micro Computer (SMCI)	US High Volatility	2023-05	+112.4%	+112.4%
Palantir (PLTR)	US High Volatility	2023-05	+89.8%	+89.8%
Marvell (MRVL)	US High Volatility	2023-05	+48.2%	+48.2%

Anthropic releases a 100K context window. Long documents, contracts, and code repositories become model inputs, and long context begins to shift from a research metric to an enterprise product selling point. Anthropic: 100K Context Windows
NVIDIA earnings turn AI expectations into GPU orders. In the first quarter of fiscal 2024, the data center business and next-quarter guidance significantly exceeded market expectations, and the market began to price AI along compute, networking, and data centers. NVIDIA: FY2024 Q1 Earnings

Judgment for this month: The upstream infrastructure can receive real orders first, while the revenue, renewal, and profit of the application layer will come much later; the longer the context, the more important the source, permissions, and verification become.

June: APIs Start Calling Functions (SOUN +49.2%)

Market Notes: US stocks continue to be priced around the “selling shovels” narrative, with Nvidia, Super Micro Computer, and networking equipment stocks strengthening; A-shares see rotation among computing power, optical modules, and servers, with pure-concept stocks like C3.ai showing noticeably higher volatility than the leading names.

Monster Stock Cumulative Table (as of 2023-06, month-end adjusted closing price compounded; Nvidia as the leading benchmark)

Target (Code)	Type	Inclusion Month	Monthly Change	Cumulative Since Inclusion
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	+11.8%	+213.6%
C3.ai (AI)	US Stock High Volatility	2023-01	-8.9%	+226.0%
BigBear.ai (BBAI)	US Stock High Volatility	2023-01	+6.3%	+248.7%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	-8.0%	+125.5%
SoundHound AI (SOUN)	US Stock High Volatility	2023-02	+49.2%	+128.9%
Eoptolink Technology (300502.SZ)	A-Share High Volatility	2023-02	-23.6%	+126.4%
Cambricon Technologies (688256.SS)	A-Share High Volatility	2023-02	-12.8%	+95.4%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	-9.2%	+138.4%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	-13.2%	+117.4%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	-6.3%	+23.3%
Super Micro Computer (SMCI)	US Stock High Volatility	2023-05	+11.3%	+136.4%
Palantir (PLTR)	US Stock High Volatility	2023-05	+4.2%	+97.8%
Marvell (MRVL)	US Stock High Volatility	2023-05	+2.2%	+51.5%

OpenAI adds function calling, longer context, and lower pricing to the API. Developers can have the model output function parameters according to a JSON Schema, making the connection between the model and external tools more orchestrable. OpenAI: Function calling and other API updates

Judgment for this month: The key to tool calling is not whether the model says “I can help you,” but whether it can reliably produce structured results that are executable and verifiable.

July: Open Weights and Closed-Source Models Progress Together (SMCI +32.5%)

Market Record: US-listed AI leaders have shifted from one-sided rallies to high-level consolidation, while A-share and Hong Kong AI trading relies more on thematic rotation; NVIDIA, AMD, optical module and server stocks remain the core, and no new “monster stocks” capable of consistently decoupling from the sector have yet emerged.

Monster Stock Cumulative Table (as of 2023-07, month-end adjusted closing composite; NVIDIA included as benchmark leader)

Target (Code)	Type	Inclusion Month	Monthly Change	Cumulative Since Inclusion
NVIDIA (NVDA)	Core Bellwether	2022-11	+10.5%	+246.5%
C3.ai (AI)	US Stock High Volatility	2023-01	+15.3%	+275.9%
BigBear.ai (BBAI)	US Stock High Volatility	2023-01	-14.5%	+198.1%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	-2.3%	+120.3%
SoundHound AI (SOUN)	US Stock High Volatility	2023-02	-48.8%	+17.2%
Eoptolink (300502.SZ)	A-Share High Volatility	2023-02	-12.2%	+98.8%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	-3.2%	+89.1%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	-1.9%	+133.9%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	-10.5%	+94.6%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	-2.3%	+20.4%
Super Micro Computer (SMCI)	US Stock High Volatility	2023-05	+32.5%	+213.2%
Palantir (PLTR)	US Stock High Volatility	2023-05	+29.4%	+155.9%
Marvell (MRVL)	US Stock High Volatility	2023-05	+9.1%	+65.2%

Meta and Microsoft release Llama 2. Model weights, starter code, and commercial usage terms are opened up more broadly, enabling developers to build their own applications either in the cloud or on-premises. Meta: Llama 2
Anthropic releases Claude 2. Longer input, coding, and reasoning capabilities extend the closed-source model competition beyond chat experiences to long documents and developer APIs. Anthropic: Claude 2

This Month’s Takeaway: “Open source” is beginning to become an industry slogan, but what truly affects deployment is licensing, weights, VRAM, inference speed, and maintenance responsibility.

August: Code Llama Seizes the Code Entry Point (000158.SZ +26.2%)

Market Notes: Nvidia’s earnings report is expected to continue supporting compute-related stocks, but high-valuation AI small-caps are beginning to diverge; Nvidia and Super Micro Computer are outperforming most software concept stocks, as the market shifts from “anything AI-related goes up” to seeking real orders.

Monster Stocks Cumulative Table (as of 2023-08, month-end adjusted close composite; NVIDIA as the leading benchmark)

Target (Code)	Type	Listed Month	This Month’s Change	Cumulative Since Listing
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	+5.6%	+265.9%
C3.ai (AI)	US Stock High Volatility	2023-01	-26.1%	+177.8%
BigBear.ai (BBAI)	US Stock High Volatility	2023-01	-18.4%	+143.3%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	+6.0%	+133.6%
SoundHound AI (SOUN)	US Stock High Volatility	2023-02	+8.2%	+26.8%
Eoptolink (300502.SZ)	A-Share High Volatility	2023-02	+0.9%	+100.6%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	-21.9%	+47.7%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	-10.1%	+110.2%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	+1.0%	+96.5%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	+1.4%	+22.1%
Super Micro Computer (SMCI)	US Stock High Volatility	2023-05	-16.7%	+160.9%
Palantir (PLTR)	US Stock High Volatility	2023-05	-24.5%	+93.2%
Marvell (MRVL)	US Stock High Volatility	2023-05	-10.6%	+47.7%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	+26.2%	+26.2%

Meta releases Code Llama. Open-source models begin to specifically target code generation, completion, and local development scenarios. Meta: Code Llama

This month’s verdict: coding formed its paid boundary first because coding tasks have testing, submission, and delivery of results, making the value much easier to verify than general chat.

September: ChatGPT Goes Multimodal (000158.SZ +21.9%)

Market Records: Long-end U.S. Treasury yields rose, and AI leaders pulled back from their highs, though Nvidia remained the relatively strong core name; the A-share application sector cooled off, with no new sustained “demon stocks” emerging.

Monster Stocks Cumulative Table (as of 2023-09, month-end adjusted close compounded; NVIDIA included as benchmark leader)

Asset (Code)	Type	Inclusion Month	Monthly Change	Cumulative Since Inclusion
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	-11.9%	+222.4%
C3.ai (AI)	US Stock High Volatility	2023-01	-17.7%	+128.6%
BigBear.ai (BBAI)	US Stock High Volatility	2023-01	-7.9%	+124.0%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	-17.8%	+92.0%
SoundHound AI (SOUN)	US Stock High Volatility	2023-02	-20.2%	+1.2%
Eoptolink Technology (300502.SZ)	A-Share High Volatility	2023-02	-31.8%	+36.8%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	-13.5%	+27.8%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	-25.3%	+57.0%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	-24.2%	+49.0%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	-9.2%	+10.9%
Super Micro Computer (SMCI)	US Stock High Volatility	2023-05	-0.3%	+160.1%
Palantir (PLTR)	US Stock High Volatility	2023-05	+6.8%	+106.4%
Marvell (MRVL)	US Stock High Volatility	2023-05	-7.1%	+37.2%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	+21.9%	+53.8%

OpenAI brings image understanding, voice, and image-based conversation to ChatGPT. Text models are beginning to become the entry point for multimodal work. OpenAI: ChatGPT image and voice updates

This Month’s Takeaway: The objects models handle have expanded from text to screenshots, meeting notes, and documents, but understanding materials does not mean knowing whether they are real.

October: Image Generation Enters the Dialogue (300502.SZ +47.5%)

Market Notes: U.S. tech stocks continue to be weighed down by interest rate expectations and export control concerns, with AI leaders trading sideways to weaker. In the A-share market, computing power and application stocks are taking turns rebounding; the rally looks more like a short-term thematic play, and no new global AI “dark horse” has emerged yet.

Monster Stocks Cumulative Table (as of 2023-10, month-end closing price with adjusted composite; NVIDIA as the leading reference)

Target (Code)	Type	Listing Month	Monthly Change	Cumulative Since Listing
Nvidia (NVDA)	Core Leader Benchmark	2022-11	-6.3%	+202.1%
C3.ai (AI)	US Stock – High Volatility	2023-01	-4.4%	+118.6%
BigBear.ai (BBAI)	US Stock – High Volatility	2023-01	-15.9%	+88.4%
Kunlun Tech (300418.SZ)	A-Share – High Volatility	2023-01	+2.6%	+97.0%
SoundHound AI (SOUN)	US Stock – High Volatility	2023-02	-20.9%	-20.0%
Eoptolink (300502.SZ)	A-Share – High Volatility	2023-02	+47.5%	+101.8%
Cambricon (688256.SS)	A-Share – High Volatility	2023-02	+39.7%	+78.5%
Foxconn Industrial Internet (601138.SS)	A-Share – High Volatility	2023-02	+2.6%	+61.1%
Innolight (300308.SZ)	A-Share – High Volatility	2023-03	+8.2%	+61.2%
Victory Giant Technology (300476.SZ)	A-Share – High Volatility	2023-04	-5.1%	+5.2%
Super Micro Computer (SMCI)	US Stock – High Volatility	2023-05	-12.7%	+127.1%
Palantir (PLTR)	US Stock – High Volatility	2023-05	-7.5%	+90.9%
Marvell (MRVL)	US Stock – High Volatility	2023-05	-12.7%	+19.8%
Changshan Beiming (000158.SZ)	A-Share – High Volatility	2023-08	-11.0%	+36.9%

DALL·E 3 enters ChatGPT Plus and Enterprise. Image generation moves from a standalone prompting tool into a conversational editing workflow, and OpenAI begins experimenting with a provenance classifier to identify generated content. OpenAI: DALL·E 3

Judgment of this month: The competition in AIGC is shifting from one-shot generation to iteration, copyright, provenance, and credibility.

November: DevDay and Governance Crisis (PLTR +35.5%)

Market Records: The leading U.S. AI stocks have regained strength amid earnings reports and the OpenAI governance crisis, with Microsoft, Nvidia, and AMD offering higher certainty than small-cap concept stocks; A-share AI trading has warmed up, but no new monster stocks have emerged that can deviate from fundamentals over the long term.

Monster Stocks Cumulative Table (as of 2023-11, month-end adjusted closing price composite; NVIDIA as the leading benchmark)

Target (Code)	Type	Listed Month	Monthly Change	Cumulative Since Listing
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	+14.7%	+246.5%
C3.ai (AI)	US High Volatility	2023-01	+19.3%	+160.7%
BigBear.ai (BBAI)	US High Volatility	2023-01	+33.9%	+152.3%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	+15.6%	+127.7%
SoundHound AI (SOUN)	US High Volatility	2023-02	+34.6%	+7.7%
Eoptolink (300502.SZ)	A-Share High Volatility	2023-02	+6.5%	+114.9%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	-9.9%	+60.8%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	+0.1%	+61.3%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	+19.0%	+91.8%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	-4.2%	+0.8%
Super Micro Computer (SMCI)	US High Volatility	2023-05	+14.2%	+159.3%
Palantir (PLTR)	US High Volatility	2023-05	+35.5%	+158.6%
Marvell (MRVL)	US High Volatility	2023-05	+18.0%	+41.4%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	-8.4%	+25.4%
Arm (ARM)	US High Volatility	2023-11	+24.8%	+24.8%

OpenAI DevDay releases GPT-4 Turbo, Assistants API, vision capabilities, and lower prices. API, retrieval, function calling, and the assistants framework further embed models into applications. OpenAI: New Models and Developer Products Announced at DevDay
OpenAI’s board abruptly fires and later reinstates Sam Altman. A governance crisis within a week makes model-company governance, shareholder relations, and talent dependency part of the industry’s reality. OpenAI: Sam Altman Returns as CEO, OpenAI Has a New Initial Board

Judgment for this month: Beyond the risks associated with the model company’s products, there are also risks related to the board of directors, core talent, capital, and partners.

December: Gemini Brings Multimodality into Products (BBAI +25.9%)

Market Notes: Leading US AI stocks are approaching their yearly highs, with NVIDIA emerging as the clearest “selling picks and shovels” mainline for the year; the A-share and Hong Kong stock markets remain structurally driven, where servers and optical modules more easily attract capital than pure application concepts.

Monster Stock Cumulative Table (as of 2023-12, month-end adjusted closing compound; NVIDIA as the leading reference)

Target (Code)	Type	Listed Month	Monthly Change	Cumulative Since Listing
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	+5.9%	+266.9%
C3.ai (AI)	US High Volatility	2023-01	-1.4%	+157.1%
BigBear.ai (BBAI)	US High Volatility	2023-01	+25.9%	+217.6%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	-20.7%	+80.6%
SoundHound AI (SOUN)	US High Volatility	2023-02	-0.9%	+6.7%
Eoptolink (300502.SZ)	A-Share High Volatility	2023-02	-14.0%	+84.8%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	-18.5%	+31.1%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	-12.0%	+41.9%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	-9.0%	+74.6%
Victory Giant (300476.SZ)	A-Share High Volatility	2023-04	-19.6%	-18.9%
Super Micro Computer (SMCI)	US High Volatility	2023-05	+3.9%	+169.5%
Palantir (PLTR)	US High Volatility	2023-05	-14.4%	+121.4%
Marvell (MRVL)	US High Volatility	2023-05	+8.2%	+53.0%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	-11.7%	+10.7%
Arm (ARM)	US High Volatility	2023-11	+22.2%	+52.5%

Google releases Gemini 1.0. Gemini is tiered from the start into Ultra, Pro, and Nano, covering cloud, applications, and on-device, with a strong emphasis on native multimodality. Google: Introducing Gemini

This Month’s Assessment: Model competition is no longer limited to a single web chat entry point, but is simultaneously fighting for search, mobile, cloud, and developer tools.

2024: Multimodality, Reasoning, and Agents Enter Products

January: GPT Store Attempts to Distribute Assistants (SMCI +86.3%)

Market Records: The beginning of 2024 was still driven by US semiconductors, with NVIDIA, AMD, Broadcom, Arm, and Super Micro Computer becoming high-beta targets; A-shares AI first rallied strongly then diverged, with no unified monster stock across the entire market yet.

Monster Stocks Cumulative Table (as of 2024-01, month-end adjusted closing composite; Nvidia as the bellwether benchmark)

Target (Code)	Type	Inclusion Month	Monthly Change	Cumulative Since Inclusion
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	+24.2%	+355.7%
C3.ai (AI)	US Stock High Volatility	2023-01	-13.7%	+121.9%
BigBear.ai (BBAI)	US Stock High Volatility	2023-01	-24.3%	+140.5%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	+36.1%	+145.8%
SoundHound AI (SOUN)	US Stock High Volatility	2023-02	-21.7%	-16.4%
Eoptolink Technology (300502.SZ)	A-Share High Volatility	2023-02	+41.6%	+161.7%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	+53.3%	+100.9%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	+40.4%	+99.3%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	+50.9%	+163.4%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	+46.4%	+18.7%
Super Micro Computer (SMCI)	US Stock High Volatility	2023-05	+86.3%	+402.0%
Palantir (PLTR)	US Stock High Volatility	2023-05	-6.3%	+107.4%
Marvell (MRVL)	US Stock High Volatility	2023-05	+12.4%	+71.9%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	+8.7%	+20.4%
Arm (ARM)	US Stock High Volatility	2023-11	-6.0%	+43.4%

OpenAI launches GPT Store. Users can publish and discover custom GPTs, as model platforms begin experimenting with combining prompts, knowledge bases, and tools into distributable software units. OpenAI: Introducing the GPT Store

Judgment of the Month: The challenges of platformization have shifted from “can we build an assistant” to discovery, distribution, permissions, and sustained usage.

February: Sora and the Long Context Race (SOUN +347.0%)

Market Notes: Nvidia’s earnings reports and AI server orders have pushed the market into an acceleration phase. High-volatility targets such as Super Micro Computer, Arm, and SoundHound have taken turns becoming “meme stocks”; A-share optical modules, servers, and computing power leasing have also shown notable activity.

Monster Stocks Cumulative Table (as of 2024-02, month-end adjusted closing composite; NVIDIA as the leading benchmark)

Asset (Code)	Type	Listed Month	Monthly Change	Cumulative Since Listing
NVIDIA (NVDA)	Core Benchmark	2022-11	+28.6%	+486.0%
C3.ai (AI)	US Stock High Volatility	2023-01	+49.2%	+231.0%
BigBear.ai (BBAI)	US Stock High Volatility	2023-01	+107.4%	+398.7%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	-1.3%	+142.6%
SoundHound AI (SOUN)	US Stock High Volatility	2023-02	+347.0%	+273.6%
Eoptolink (300502.SZ)	A-Share High Volatility	2023-02	+11.6%	+192.1%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	+2.8%	+106.5%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	+21.8%	+142.7%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	+0.9%	+165.8%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	+11.5%	+32.3%
Super Micro Computer (SMCI)	US Stock High Volatility	2023-05	+63.5%	+720.8%
Palantir (PLTR)	US Stock High Volatility	2023-05	+55.9%	+223.4%
Marvell (MRVL)	US Stock High Volatility	2023-05	+5.8%	+81.9%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	-6.5%	+12.6%
Arm (ARM)	US Stock High Volatility	2023-11	+99.6%	+186.1%

OpenAI releases research preview of Sora. Text-to-video brings the generative model competition to temporal consistency, physical world simulation, and video source verification. OpenAI: Sora
Google releases Gemini 1.5, emphasizing long context and MoE architecture. The model begins handling longer videos, codebases, and multi-document inputs. Google: Gemini 1.5

Judgment of the Month: The next threshold for multimodality is not generating a single beautiful frame, but maintaining coherence over long durations and within long contexts.

March: Models Begin to Stratify by Capability (300502.SZ +29.1%)

Market Records: The GTC and NVIDIA’s new-generation computing power narrative continue to push semiconductors higher, with NVIDIA, Super Micro Computer, Arm, and the server chain seeing high-level position rotation; highly elastic A-share targets such as Cambricon, Industrial Fulian, and Zhongji Xudong experienced significant volatility.

Monster Stocks Cumulative Table (As of 2024-03, Month-End Adjusted Close Compound; NVIDIA as Benchmark Leader)

Target (Code)	Type	Inclusion Month	Monthly Change	Cumulative Since Inclusion
NVIDIA (NVDA)	Core Bellwether	2022-11	+14.2%	+569.2%
C3.ai (AI)	US High Volatility	2023-01	-26.8%	+142.3%
BigBear.ai (BBAI)	US High Volatility	2023-01	-39.0%	+204.2%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	+0.3%	+143.3%
SoundHound AI (SOUN)	US High Volatility	2023-02	-20.6%	+196.7%
Eoptolink (300502.SZ)	A-Share High Volatility	2023-02	+29.1%	+277.1%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	-1.2%	+104.1%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	+7.6%	+161.2%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	+19.0%	+216.3%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	+22.5%	+62.1%
Super Micro Computer (SMCI)	US High Volatility	2023-05	+16.6%	+857.0%
Palantir (PLTR)	US High Volatility	2023-05	-8.3%	+196.6%
Marvell (MRVL)	US High Volatility	2023-05	-1.1%	+79.9%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	-4.5%	+7.5%
Arm (ARM)	US High Volatility	2023-11	-11.4%	+153.5%

Anthropic releases the Claude 3 family. Opus, Sonnet, and Haiku cover the enterprise API with different capabilities, speeds, and costs; model selection begins to turn into a routing problem. Anthropic: Claude 3 family

Verdict for this month: “The strongest model” is no longer the only answer; latency, price, and task difficulty are equally decisive in determining whether a product is usable.

April: Llama 3 Expands the Open Ecosystem (300308.SZ -39.7%)

Market Recap: Rising interest rates and earnings expectations triggered a pullback in US AI stocks, with Super Micro Computer and some high-valuation small caps seeing steeper declines; the A-share AI theme cooled in sync, with the speculative rally temporarily fading.

Meme Stock Cumulative Table (as of 2024-04, month-end closing price with adjusted composite; NVIDIA as the leading benchmark)

Target (Code)	Type	Listed Month	Monthly Change	Cumulative Since Listing
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	-4.4%	+539.8%
C3.ai (AI)	US High Volatility	2023-01	-16.8%	+101.6%
BigBear.ai (BBAI)	US High Volatility	2023-01	-19.0%	+146.4%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	-12.7%	+112.4%
SoundHound AI (SOUN)	US High Volatility	2023-02	-28.0%	+113.6%
Eoptolink (300502.SZ)	A-Share High Volatility	2023-02	+0.3%	+278.2%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	+2.0%	+108.1%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	-6.5%	+144.2%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	-39.7%	+90.7%
Victory Giant (300476.SZ)	A-Share High Volatility	2023-04	-7.9%	+49.3%
Super Micro Computer (SMCI)	US High Volatility	2023-05	-15.0%	+713.5%
Palantir (PLTR)	US High Volatility	2023-05	-4.5%	+183.2%
Marvell (MRVL)	US High Volatility	2023-05	-6.9%	+67.5%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	-5.0%	+2.1%
Arm (ARM)	US High Volatility	2023-11	-19.0%	+105.3%

Meta Releases Llama 3. The 8B and 70B models are open-sourced to the community, with expanded training data scale, tooling, and cloud partners. Meta: Llama 3

This month’s verdict: Open-weight models are shifting from research alternatives to a complete ecosystem, where deployers can trade cost and data boundaries for control.

May: GPT-4o Enables Real-Time Interaction (NVDA +26.9%)

Market Records: NVIDIA’s earnings report once again confirms AI capital expenditures, with related stocks such as NVIDIA, Dell, Super Micro Computer, and TSMC strengthening; A-share optical module and server chains followed, with funds shifting back toward “hardware monster stocks” where orders can be seen.

Monster Stocks Cumulative Table (as of 2024-05, month-end adjusted close composite; NVIDIA included as bellwether reference)

Target (Code)	Type	Inclusion Month	Monthly Change	Cumulative Since Inclusion
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	+26.9%	+711.9%
C3.ai (AI)	US Stock High Volatility	2023-01	+31.2%	+164.5%
BigBear.ai (BBAI)	US Stock High Volatility	2023-01	-9.6%	+122.8%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	-7.5%	+96.5%
SoundHound AI (SOUN)	US Stock High Volatility	2023-02	+19.1%	+154.4%
Eoptolink Technology (300502.SZ)	A-Share High Volatility	2023-02	+21.7%	+360.3%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	+13.7%	+136.7%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	+19.6%	+192.1%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	+23.1%	+134.8%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	+18.7%	+77.2%
Super Micro Computer (SMCI)	US Stock High Volatility	2023-05	-8.7%	+642.7%
Palantir (PLTR)	US Stock High Volatility	2023-05	-1.3%	+179.5%
Marvell (MRVL)	US Stock High Volatility	2023-05	+4.4%	+74.9%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	-8.6%	-6.7%
Arm (ARM)	US Stock High Volatility	2023-11	+19.1%	+144.6%

OpenAI released GPT-4o. Text, vision, and audio capabilities were unified into a single faster model, with real-time voice interaction becoming a product focus. OpenAI: GPT-4o

Judgment for this month: Multimodal is no longer just about uploading files; real-time interaction is beginning to compete for mobile, customer service, and office entry points.

June: AI Enters Devices and Workflows (ARM +35.8%)

Market Records: Nvidia briefly became the world’s most valuable company by market capitalization in June, with stock splits and index funds further amplifying the attention; in A-shares, computing power chain stocks such as Zhongji Innolight, Eoptolink, and Industrial Fulian outperformed most application stocks, and Nvidia was the most typical AI leader during this phase.

Monster Stocks Cumulative Table (As of 2024-06, Month-End Adjusted Composite Close; NVIDIA as Benchmark Leader)

Target (Code)	Type	Entry Month	Monthly Change	Cumulative Since Entry
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	+12.7%	+815.0%
C3.ai (AI)	US Stock High Volatility	2023-01	-2.1%	+158.9%
BigBear.ai (BBAI)	US Stock High Volatility	2023-01	+0.7%	+124.3%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	-6.5%	+83.7%
SoundHound AI (SOUN)	US Stock High Volatility	2023-02	-21.8%	+98.9%
Eoptolink Technology (300502.SZ)	A-Share High Volatility	2023-02	-5.3%	+335.9%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	+32.9%	+214.5%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	-13.0%	+154.1%
Zhongji Innolight (300308.SZ)	A-Share High Volatility	2023-03	-6.1%	+120.5%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	+19.5%	+111.7%
Super Micro Computer (SMCI)	US Stock High Volatility	2023-05	+4.4%	+675.4%
Palantir (PLTR)	US Stock High Volatility	2023-05	+16.8%	+226.5%
Marvell (MRVL)	US Stock High Volatility	2023-05	+1.6%	+77.7%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	-0.3%	-6.9%
Arm (ARM)	US Stock High Volatility	2023-11	+35.8%	+232.1%

Apple launches Apple Intelligence. On-device models, personal context, Private Cloud Compute, and ChatGPT integration are woven into the system experience of iPhone, iPad, and Mac. Apple: Apple Intelligence
Anthropic releases Claude 3.5 Sonnet. Improvements in code, vision, and tool use push coding and Agent tasks to the center of the closed-source model competition. Anthropic: Claude 3.5 Sonnet

Judgment of the Month: AI is moving beyond “opening a website” and becoming embedded into devices and workflows, with privacy, permissions, and latency now part of the experience.

July: Large Models Shift Toward Low-Cost Routing (SOUN +28.9%)

Market Record: The market is rotating from large-cap tech stocks to small-caps and non-AI sectors, with NVIDIA and Super Micro Computer rallying first before pulling back; A-share computing power stocks are oscillating at high levels, and the characteristics of speculative stocks have shifted from “continuous rallies” to “high volatility and rapid rotation.”

Monsters Stock Cumulative Table (as of end of July 2024, month-end adjusted closing composite; NVIDIA as the bellwether benchmark)

Target (Code)	Type	Inception Month	MTD Change	Cumulative Since Inception
NVIDIA (NVDA)	Core Benchmark	2022-11	-5.3%	+766.5%
C3.ai (AI)	US High-Vol	2023-01	-7.6%	+139.3%
BigBear.ai (BBAI)	US High-Vol	2023-01	+0.0%	+124.3%
Kunlun Tech (300418.SZ)	A-Share High-Vol	2023-01	-9.9%	+65.5%
SoundHound AI (SOUN)	US High-Vol	2023-02	+28.9%	+156.4%
Eoptolink (300502.SZ)	A-Share High-Vol	2023-02	-6.0%	+309.7%
Cambricon (688256.SS)	A-Share High-Vol	2023-02	-2.8%	+205.7%
Foxconn Industrial Internet (601138.SS)	A-Share High-Vol	2023-02	-11.7%	+124.4%
Innolight (300308.SZ)	A-Share High-Vol	2023-03	-15.9%	+85.4%
Victory Giant Technology (300476.SZ)	A-Share High-Vol	2023-04	-12.3%	+85.7%
Super Micro Computer (SMCI)	US High-Vol	2023-05	-14.4%	+563.7%
Palantir (PLTR)	US High-Vol	2023-05	+6.2%	+246.7%
Marvell (MRVL)	US High-Vol	2023-05	-4.1%	+70.4%
Changshan Beiming (000158.SZ)	A-Share High-Vol	2023-08	+21.9%	+13.4%
Arm (ARM)	US High-Vol	2023-11	-11.9%	+192.6%

Meta releases Llama 3.1 405B. Open-weight models begin directly competing with frontier closed-source models, and provide long-context and agent reference systems. Meta: Llama 3.1
OpenAI releases GPT-4o mini. Smaller and cheaper models drive localization or batching of simple tasks, and the economics of model routing become clearer. OpenAI: GPT-4o mini

Judgment for this month: The unit economics of AI products don’t depend solely on the strongest model. Low-cost models that handle large volumes of simple tasks are what give API calls a chance to become long-term budget items.

August: AI Act Enters into Implementation (000158.SZ +69.3%)

Market Notes: The unwinding of the yen carry trade triggered a sharp selloff in global tech stocks followed by a rapid recovery, with the price swings of NVIDIA and AMD widening once again; AI stocks in the A-share market are still primarily driven by rotation, without the emergence of any new single monster stock.

Monster Stock Cumulative Table (As of 2024-08, Month-end adjusted closing composite; NVIDIA as the bellwether benchmark)

Target (Code)	Type	Inclusion Month	Monthly Change	Cumulative Since Inclusion
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	+2.0%	+783.9%
C3.ai (AI)	US High Volatility	2023-01	-12.7%	+108.9%
BigBear.ai (BBAI)	US High Volatility	2023-01	+5.3%	+136.2%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	+38.3%	+128.9%
SoundHound AI (SOUN)	US High Volatility	2023-02	-3.9%	+146.4%
Eoptolink (300502.SZ)	A-Share High Volatility	2023-02	+38.6%	+467.9%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	+12.6%	+244.2%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	+23.1%	+176.2%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	+42.2%	+163.6%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	+17.4%	+118.0%
Super Micro Computer (SMCI)	US High Volatility	2023-05	-37.6%	+314.2%
Palantir (PLTR)	US High Volatility	2023-05	+17.1%	+306.0%
Marvell (MRVL)	US High Volatility	2023-05	+13.8%	+93.9%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	+69.3%	+92.0%
Arm (ARM)	US High Volatility	2023-11	-7.8%	+169.8%

EU AI Act takes effect. Risk classification, transparency, and general-purpose AI model obligations begin entering the formal implementation phase. EUR-Lex: Regulation (EU) 2024/1689

This month’s verdict: Regulation is no longer an abstract discussion before launch; it will gradually make its way into model documentation, deployment processes, and enterprise procurement checklists.

September: A new curve emerges for reasoning models (000158.SZ +180.8%)

Market Notes: Rate cut expectations support a rebound in U.S. tech stocks, but capital is starting to diversify from high-valuation chips to software and platforms; Nvidia, Arm, and Super Micro Computer remain representative of high volatility, while A-share AI is in repeated bottom-building.

Monster Stocks Cumulative Table (as of 2024-09, month-end adjusted closing composite; NVIDIA as benchmark leader)

Ticker (Code)	Type	Listed Month	Monthly Change	Cumulative Since Listing
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	+1.7%	+798.9%
C3.ai (AI)	US Stock High Volatility	2023-01	+3.8%	+116.8%
BigBear.ai (BBAI)	US Stock High Volatility	2023-01	-8.2%	+116.8%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	+8.5%	+148.4%
SoundHound AI (SOUN)	US Stock High Volatility	2023-02	-4.7%	+134.8%
Eoptolink (300502.SZ)	A-Share High Volatility	2023-02	+0.2%	+469.0%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	+56.9%	+440.1%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	-3.7%	+166.0%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	-7.6%	+143.6%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	+13.4%	+147.2%
Super Micro Computer (SMCI)	US Stock High Volatility	2023-05	-4.9%	+293.9%
Palantir (PLTR)	US Stock High Volatility	2023-05	+18.2%	+379.9%
Marvell (MRVL)	US Stock High Volatility	2023-05	-5.4%	+83.4%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	+180.8%	+439.2%
Arm (ARM)	US Stock High Volatility	2023-11	+7.6%	+190.3%

OpenAI released o1-preview and o1-mini. Reasoning models shift more computation into the inference stage, where math, code, and complex tasks begin to be priced based on “how much they think.” OpenAI: o1-preview and o1-mini

This month’s verdict: Beyond “bigger,” a capability curve of “thinking longer” has emerged. Waiting, token, and review costs all need to be calculated together.

October: Agents begin operating computers and searching (SMCI -30.1%)

Market Recap: Leading U.S. AI stocks diverged around earnings reports, with Nvidia remaining relatively resilient, while Super Micro Computer and small-cap AI stocks saw larger volatility. In A-shares, capital rotated toward new themes such as computing power, satellite communications, and robotics, with the performance of speculative “demon stocks” clearly becoming shorter in cycle.

Monster Stocks Cumulative Table (as of 2024-10, month-end adjusted close compounded; NVIDIA as the leading stock benchmark)

Target (Code)	Type	Inclusion Month	Monthly Change	Cumulative Since Inclusion
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	+9.3%	+882.5%
C3.ai (AI)	US Stock High Volatility	2023-01	+1.7%	+120.5%
BigBear.ai (BBAI)	US Stock High Volatility	2023-01	+8.9%	+136.1%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	+13.2%	+181.1%
SoundHound AI (SOUN)	US Stock High Volatility	2023-02	+7.9%	+153.4%
Eoptolink Technology (300502.SZ)	A-Share High Volatility	2023-02	-11.7%	+402.4%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	+23.7%	+568.1%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	-8.5%	+143.4%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	-11.4%	+115.8%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	-9.9%	+122.7%
Super Micro Computer (SMCI)	US Stock High Volatility	2023-05	-30.1%	+175.3%
Palantir (PLTR)	US Stock High Volatility	2023-05	+11.7%	+436.1%
Marvell (MRVL)	US Stock High Volatility	2023-05	+11.2%	+104.0%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	-29.5%	+280.2%
Arm (ARM)	US Stock High Volatility	2023-11	-1.2%	+186.8%

Anthropic releases computer use capability. Claude can now view the screen, move the mouse, click, and type—agents evolve from calling APIs to operating general-purpose software. Anthropic: computer use
OpenAI releases ChatGPT Search. The chat entry point now directly returns real-time web answers with links, further narrowing the boundary between search and Q&A. OpenAI: ChatGPT search

Judgment for this month: The value of an Agent comes from execution, but the stronger the execution capability, the less the permissions, prompt injection, error rollback, and responsibility boundaries can be omitted.

November: MCP Connects Tools and Data (SOUN +85.1%)

Market Record: After the U.S. election, AI themes heated up alongside software, autonomous driving, and robotics sectors. Targets such as NVIDIA, Palantir, and SoundHound strengthened but experienced sharp divergence; Hong Kong tech stocks rebounded in tandem, though a stable basket of “AI monster stocks” has yet to form.

Monster Stocks Cumulative Table (as of 2024-11, month-end adjusted closing composite; NVIDIA as the leading benchmark)

Target (Code)	Type	Inclusion Month	Monthly Change	Cumulative Since Inclusion
NVIDIA (NVDA)	Core Benchmark	2022-11	+4.1%	+922.8%
C3.ai (AI)	US High Volatility	2023-01	+51.0%	+233.0%
BigBear.ai (BBAI)	US High Volatility	2023-01	+44.0%	+240.0%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	-16.5%	+134.7%
SoundHound AI (SOUN)	US High Volatility	2023-02	+85.1%	+369.0%
Eoptolink (300502.SZ)	A-Share High Volatility	2023-02	+0.5%	+405.0%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	+17.3%	+683.7%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	-3.1%	+135.8%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	-2.5%	+110.4%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	+3.8%	+131.2%
Super Micro Computer (SMCI)	US High Volatility	2023-05	+12.1%	+208.6%
Palantir (PLTR)	US High Volatility	2023-05	+61.4%	+765.3%
Marvell (MRVL)	US High Volatility	2023-05	+15.7%	+136.0%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	-22.9%	+193.1%
Arm (ARM)	US High Volatility	2023-11	-5.0%	+172.5%

Anthropic open-sources Model Context Protocol. MCP attempts to connect models with external tools and data sources through a universal protocol, signaling that tool integration is evolving from single-company implementations toward an ecosystem-wide protocol. Anthropic: Model Context Protocol

Judgment of the Month: The bottleneck for Agents is gradually shifting from the model itself to tool protocols, context management, and permission systems.

December: Sora Goes Live, Gemini Enters the Agent Era (SOUN +113.1%)

Market Records: By year-end, AI trading expanded from chips to software, data, and robotics. Nvidia experienced volatile trading at high levels, while high-beta names like Palantir and SoundHound behaved more like meme stocks; the A-share computing power chain failed to replicate the broad rally seen at the beginning of the year.

Monstrous Stocks Cumulative Table (as of the end of 2024-12, month-end adjusted closing price compound; Nvidia as the leading comparison)

Target (Code)	Type	Month Added	This Month Change	Cumulative Since Added
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	-2.9%	+893.1%
C3.ai (AI)	US Stock High Volatility	2023-01	-7.4%	+208.3%
BigBear.ai (BBAI)	US Stock High Volatility	2023-01	+94.3%	+560.7%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	-4.1%	+125.1%
SoundHound AI (SOUN)	US Stock High Volatility	2023-02	+113.1%	+899.5%
Eoptolink Technology (300502.SZ)	A-Share High Volatility	2023-02	+8.8%	+449.4%
Cambricon Technologies (688256.SS)	A-Share High Volatility	2023-02	-13.1%	+581.0%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	-0.2%	+135.3%
Innolight Technology (300308.SZ)	A-Share High Volatility	2023-03	-7.0%	+95.7%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	+26.5%	+192.5%
Super Micro Computer (SMCI)	US Stock High Volatility	2023-05	-6.6%	+188.3%
Palantir (PLTR)	US Stock High Volatility	2023-05	+12.7%	+875.1%
Marvell (MRVL)	US Stock High Volatility	2023-05	+19.2%	+181.3%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	-11.7%	+158.8%
Arm (ARM)	US Stock High Volatility	2023-11	-8.1%	+150.4%

OpenAI officially released Sora. Video generation moved from research preview to a subscription product, with watermarking and provenance identification measures enabled by default. OpenAI: Sora is here
Google released Gemini 2.0, emphasizing the Agentic Era. Multimodal output, tool use, and real-time interaction became the next-generation product direction. Google: Gemini 2.0

This month’s takeaway: The main storyline of 2024 shifts from “what a model can generate” to “whether a model can complete a sequence of actions in real-world environments.”

2025: Reasoning Costs Drop, Coding Becomes the Agent Proving Ground

January: R1 Recalculation Costs, Stargate Increases Infrastructure Investment (688158.SS +155.2%)

Market Record: On January 27, the DeepSeek shock caused Nvidia to drop nearly 17% in a single day, resulting in a rare massive market cap pullback; the U.S. stock computing power chain was sold off first, while A-share concept stocks first saw abnormal movements and then spread after the holiday, with Zhejiang Eastern, Meig Smart, Every Daily Interactive, UCloud, and others becoming representatives of high volatility. This month also saw a split行情 of “selling computing power overseas” and “buying applications domestically.” Reuters: DeepSeek Triggers AI Stock Sell-off

Monster Stock Cumulative Table (as of 2025-01, month-end adjusted closing composite; NVIDIA as the leading benchmark)

Asset (Code)	Type	Listed Month	Monthly Change	Cumulative Since Listing
NVIDIA (NVDA)	Core Bellwether	2022-11	-10.6%	+787.8%
C3.ai (AI)	US High-Volatility	2023-01	-8.9%	+180.9%
BigBear.ai (BBAI)	US High-Volatility	2023-01	-4.7%	+529.6%
Kunlun Tech (300418.SZ)	A-Share High-Volatility	2023-01	+0.3%	+125.8%
SoundHound AI (SOUN)	US High-Volatility	2023-02	-28.7%	+612.6%
Eoptolink (300502.SZ)	A-Share High-Volatility	2023-02	-23.6%	+319.7%
Cambricon (688256.SS)	A-Share High-Volatility	2023-02	+28.6%	+775.8%
Foxconn Industrial Internet (601138.SS)	A-Share High-Volatility	2023-02	-1.2%	+132.5%
Innolight (300308.SZ)	A-Share High-Volatility	2023-03	-12.1%	+72.0%
Victory Giant Technology (300476.SZ)	A-Share High-Volatility	2023-04	-4.1%	+180.5%
Super Micro Computer (SMCI)	US High-Volatility	2023-05	-6.4%	+169.8%
Palantir (PLTR)	US High-Volatility	2023-05	+9.1%	+963.9%
Marvell (MRVL)	US High-Volatility	2023-05	+2.2%	+187.5%
Changshan Beiming (000158.SZ)	A-Share High-Volatility	2023-08	+41.6%	+266.5%
Arm (ARM)	US High-Volatility	2023-11	+29.3%	+223.8%
Zhejiang Orient Holdings (600120.SS)	A-Share Thematic High-Volatility	2025-01	+43.0%	+43.0%
MeiG Smart (002881.SZ)	A-Share Thematic High-Volatility	2025-01	+40.9%	+40.9%
UCloud (688158.SS)	A-Share Thematic High-Volatility	2025-01	+155.2%	+155.2%
QingCloud Technology (688316.SS)	A-Share Thematic High-Volatility	2025-01	+139.2%	+139.2%
Talkweb Information (002261.SZ)	A-Share Thematic High-Volatility	2025-01	+98.5%	+98.5%

DeepSeek released DeepSeek-R1 and open-sourced its weights and paper. The open reasoning model brought frontier capabilities, training methodology, and unit economics all into the global market spotlight at once. DeepSeek: R1 Release
The U.S. announced the Stargate project. Partners including OpenAI, SoftBank, and Oracle unveiled a large-scale AI infrastructure plan, with the capex narrative continuing to concentrate around data centers. OpenAI: Stargate Project

This month’s verdict: DeepSeek forces the market to recompute the cost per unit of capability, while Stargate demonstrates that total demand, total investment, and the cost per inference can all rise simultaneously.

February: Research, reasoning, and coding extending task chains (300476.SZ +58.7%)

Market Notes: The DeepSeek theme continues to spread in the A-share market. After Zhejiang Eastern continues to hit the daily limit up/down, Magic Smart, UCloud, QingCloud Technology, and Talkweb Information take turns. The U.S. stock market is recovering from panic, with NVIDIA and cloud vendors returning to the logic of capital expenditure. A-share “demon stocks” and U.S. stock leaders are running at two different paces. Risk warning regarding Zhejiang Eastern and Magic Smart

Monster Stocks Cumulative Table (As of 2025-02, month-end adjusted closing composite; NVIDIA included as the bellwether reference)

Target (Code)	Type	Inclusion Month	Monthly Change	Cumulative Since Inclusion
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	+4.0%	+823.3%
C3.ai (AI)	US Stock High Volatility	2023-01	-25.2%	+110.1%
BigBear.ai (BBAI)	US Stock High Volatility	2023-01	+21.7%	+666.3%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	-6.9%	+110.2%
SoundHound AI (SOUN)	US Stock High Volatility	2023-02	-23.5%	+445.2%
Eoptolink Technology (300502.SZ)	A-Share High Volatility	2023-02	+2.2%	+329.0%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	-15.3%	+641.8%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	-6.3%	+117.9%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	-2.2%	+68.2%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	+58.7%	+345.1%
Super Micro Computer (SMCI)	US Stock High Volatility	2023-05	+45.4%	+292.3%
Palantir (PLTR)	US Stock High Volatility	2023-05	+2.9%	+994.7%
Marvell (MRVL)	US Stock High Volatility	2023-05	-18.6%	+134.0%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	-11.9%	+222.9%
Arm (ARM)	US Stock High Volatility	2023-11	-17.5%	+167.1%
Zhejiang Orient (600120.SS)	A-Share Thematic High Volatility	2025-01	-12.1%	+25.7%
Meig Smart (002881.SZ)	A-Share Thematic High Volatility	2025-01	-15.9%	+18.5%
UCloud (688158.SS)	A-Share Thematic High Volatility	2025-01	-26.3%	+88.1%
QingCloud Technology (688316.SS)	A-Share Thematic High Volatility	2025-01	-17.0%	+98.5%
Talkweb Information (002261.SZ)	A-Share Thematic High Volatility	2025-01	-14.6%	+69.5%

OpenAI released deep research. The model began independently performing multi-step web searches, reading, and synthesis, making research tasks a representative scenario for Agents. OpenAI: deep research
Anthropic released Claude 3.7 Sonnet and Claude Code. A hybrid reasoning model and a terminal coding tool emerged simultaneously, making software engineering a testing ground for long-horizon Agents. Anthropic: Claude 3.7 Sonnet and Claude Code
OpenAI released GPT-4.5 research preview. Large-scale pre-training and reasoning models formed two complementary paths, and attention began to shift toward training and serving costs beyond the model name itself. OpenAI: GPT-4.5

Judgment for this month: Reasoning models, web-connected research, and coding agents have collectively lengthened task chains and amplified the costs of context, waiting, and acceptance.

March: Web Search Becomes an API Capability (NVDA -13.2%)

Market Highlights: The market has shifted from the DeepSeek application theme to inference, computing power, and robotics. NVIDIA has regained attention around GTC; A-shares Tuowei Information, Changshan Beiming, Inspur Information, and the optical module chain have experienced high volatility, while Hong Kong tech stocks remain relatively active.

Monster Stocks Cumulative Table (as of 2025-03, month-end adjusted close compound; NVIDIA as the leader benchmark)

Target (Code)	Type	Listing Month	Monthly Change	Cumulative Since Listing
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	-13.2%	+701.5%
C3.ai (AI)	US High-Volatility	2023-01	-10.2%	+88.7%
BigBear.ai (BBAI)	US High-Volatility	2023-01	-44.6%	+324.5%
Kunlun Tech (300418.SZ)	A-Share High-Volatility	2023-01	-7.4%	+94.7%
SoundHound AI (SOUN)	US High-Volatility	2023-02	-25.0%	+308.9%
Eoptolink (300502.SZ)	A-Share High-Volatility	2023-02	-8.5%	+292.5%
Cambricon (688256.SS)	A-Share High-Volatility	2023-02	+12.9%	+737.5%
Foxconn Industrial Internet (601138.SS)	A-Share High-Volatility	2023-02	-9.1%	+98.0%
Innolight (300308.SZ)	A-Share High-Volatility	2023-03	-15.4%	+42.3%
Victory Giant (300476.SZ)	A-Share High-Volatility	2023-04	-9.3%	+303.7%
Super Micro Computer (SMCI)	US High-Volatility	2023-05	-17.4%	+224.0%
Palantir (PLTR)	US High-Volatility	2023-05	-0.6%	+988.2%
Marvell (MRVL)	US High-Volatility	2023-05	-32.9%	+57.0%
Changshan Beiming (000158.SZ)	A-Share High-Volatility	2023-08	+1.7%	+228.4%
Arm (ARM)	US High-Volatility	2023-11	-18.9%	+116.6%
Zhejiang Orient (600120.SS)	A-Share Thematic High-Volatility	2025-01	-6.6%	+17.4%
Meig Smart (002881.SZ)	A-Share Thematic High-Volatility	2025-01	+3.0%	+22.1%
UCloud (688158.SS)	A-Share Thematic High-Volatility	2025-01	-13.5%	+62.7%
QingCloud (688316.SS)	A-Share Thematic High-Volatility	2025-01	-16.7%	+65.4%
Talkweb Information (002261.SZ)	A-Share Thematic High-Volatility	2025-01	+16.4%	+97.3%

Google releases Gemini 2.5. “Thinking models” have become the core narrative of the Gemini series, with long-context and multimodal reasoning continuing to advance toward productization. Google: Gemini 2.5
Anthropic launches Web Search for the Claude API. Search capability has expanded from a standalone chat product into an API tool that developers can compose. Anthropic: Claude can now search the web

Judgment for This Month: Connectivity is no longer just a product toggle, but an orchestratable capability within the Agent architecture.

April: Model Comparison Unit Becomes a Complete Task (PLTR +40.3%)

Market Notes: Tariff and growth concerns triggered a rapid pullback in global tech stocks, with NVIDIA, AMD, Super Micro Computer, and AI infrastructure stocks falling before recovering; A-share high-level computing power stocks diverged, as the market began prioritizing “whether orders can be landed” over thematic hype.

Monster Stock Cumulative Table (as of 2025-04, month-end adjusted closing compound; NVIDIA as the leading benchmark)

Target (Code)	Type	Inclusion Month	This Month’s Change	Cumulative Since Inclusion
Nvidia (NVDA)	Core Benchmark	2022-11	+0.5%	+705.5%
C3.ai (AI)	US Stock High Volatility	2023-01	+4.6%	+97.3%
BigBear.ai (BBAI)	US Stock High Volatility	2023-01	+19.2%	+406.0%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	+4.8%	+104.0%
SoundHound AI (SOUN)	US Stock High Volatility	2023-02	+14.4%	+367.8%
Eoptolink Technology (300502.SZ)	A-Share High Volatility	2023-02	+38.2%	+442.4%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	-14.2%	+618.6%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	+4.8%	+107.6%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	+12.0%	+59.4%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	+18.2%	+377.2%
Super Micro Computer (SMCI)	US Stock High Volatility	2023-05	-7.0%	+201.4%
Palantir (PLTR)	US Stock High Volatility	2023-05	+40.3%	+1426.7%
Marvell (MRVL)	US Stock High Volatility	2023-05	-5.1%	+49.0%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	-2.2%	+221.1%
Arm (ARM)	US Stock High Volatility	2023-11	+6.8%	+131.3%
Zhejiang Orient (600120.SS)	A-Share Thematic High Volatility	2025-01	-6.2%	+10.1%
Meig Smart (002881.SZ)	A-Share Thematic High Volatility	2025-01	-11.9%	+7.5%
UCloud (688158.SS)	A-Share Thematic High Volatility	2025-01	-13.3%	+41.1%
QingCloud Technology (688316.SS)	A-Share Thematic High Volatility	2025-01	-12.7%	+44.4%
Tarchy Information (002261.SZ)	A-Share Thematic High Volatility	2025-01	-12.0%	+73.6%

Meta releases Llama 4 Scout and Maverick. The open-weight models add native multimodality and MoE architecture, with the community roadmap continuing to compete for deployment rights. Meta: Llama 4
OpenAI releases the GPT-4.1 family. Long context, instruction following, and coding become the main selling points of the API release. OpenAI: GPT-4.1
OpenAI releases o3 and o4-mini. For the first time, the reasoning models combine tools such as search, file analysis, Python, and image generation within ChatGPT, making the Agent form more complete. OpenAI: o3 and o4-mini

Judgment for this month: The unit of comparison for model releases has shifted from a single response to an entire task, as tool use, code modification, and result verification are being incorporated into the same product loop.

May: Claude 4 Competes for Long-Duration Coding Tasks (300308.SZ +56.5%)

Market recap: U.S. AI leaders rebounded on earnings and capex expectations, with NVIDIA, Broadcom, Dell, and cloud providers outperforming most application stocks; in A-shares, optical modules, PCB, and server supply chains reactivated, with speculative stock characteristics concentrated in细分 segments of the industrial chain.

Monster Stocks Cumulative Table (as of 2025-05, month-end adjusted closing composite; NVIDIA as the leading benchmark)

Target (Code)	Type	Inclusion Month	Monthly Change	Cumulative Since Inclusion
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	+24.1%	+899.6%
C3.ai (AI)	US Stock High Volatility	2023-01	+20.8%	+138.4%
BigBear.ai (BBAI)	US Stock High Volatility	2023-01	+22.0%	+517.3%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	+0.6%	+105.2%
SoundHound AI (SOUN)	US Stock High Volatility	2023-02	+8.8%	+408.9%
Eoptolink (300502.SZ)	A-Share High Volatility	2023-02	+43.9%	+680.6%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	-0.4%	+615.7%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	+12.9%	+134.3%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	+56.5%	+149.5%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	+55.4%	+641.6%
Super Micro Computer (SMCI)	US Stock High Volatility	2023-05	+25.6%	+278.5%
Palantir (PLTR)	US Stock High Volatility	2023-05	+11.3%	+1599.2%
Marvell (MRVL)	US Stock High Volatility	2023-05	+3.1%	+53.6%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	-0.2%	+220.5%
Arm (ARM)	US Stock High Volatility	2023-11	+9.2%	+152.6%
Zhejiang Orient (600120.SS)	A-Share Thematic High Volatility	2025-01	+9.4%	+20.5%
MeiG Smart (002881.SZ)	A-Share Thematic High Volatility	2025-01	+7.5%	+15.6%
UCloud (688158.SS)	A-Share Thematic High Volatility	2025-01	+16.1%	+63.8%
QingCloud Technology (688316.SS)	A-Share Thematic High Volatility	2025-01	+28.7%	+85.8%
Talkweb Information (002261.SZ)	A-Share Thematic High Volatility	2025-01	+5.7%	+83.5%

Anthropic released Claude 4. Opus 4, Sonnet 4, and Agent tool capabilities are further integrated, making long-duration coding tasks an important benchmark in frontier model competition. Anthropic: Claude 4
Anthropic released Web Search and connectivity capabilities for API. Model vendors are starting to make retrieval, file access, and external service connections into reusable developer infrastructure. Anthropic Newsroom

This Month’s Verdict: The competition in coding is no longer about completing a few lines of code, but about reading repositories, splitting tasks, modifying files, running tests, and handling failures.

June: Terminal Becomes an Agent Working Environment (601138.SS +65.1%)

Market Record: US stocks continue to trade AI capital expenditure, with NVIDIA, Broadcom, and data center suppliers maintaining strength; in A-shares, Zhongji Xudong, Xin-Yi Sheng, Shenghong Technology, and Foxconn Industrial Internet are seeing high-level expansion in the computing power chain. Optical modules and PCB are the closest sectors to “demon stocks” at this stage, but industry prosperity and individual stock prices should not be equated.

Monster Stock Cumulative Table (As of 2025-06, month-end adjusted close compounded; NVIDIA as the bellwether benchmark)

Target (Code)	Type	Listed Month	Monthly Change	Cumulative Since Listing
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	+16.9%	+1068.5%
C3.ai (AI)	US Stock High Volatility	2023-01	-7.6%	+120.3%
BigBear.ai (BBAI)	US Stock High Volatility	2023-01	+63.2%	+907.5%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	+6.7%	+119.0%
SoundHound AI (SOUN)	US Stock High Volatility	2023-02	+6.1%	+440.0%
Eoptolink (300502.SZ)	A-Share High Volatility	2023-02	+49.0%	+1063.1%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	+18.0%	+744.5%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	+65.1%	+286.9%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	+49.2%	+272.2%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	+42.9%	+959.7%
Super Micro Computer (SMCI)	US Stock High Volatility	2023-05	+22.5%	+363.7%
Palantir (PLTR)	US Stock High Volatility	2023-05	+3.4%	+1657.0%
Marvell (MRVL)	US Stock High Volatility	2023-05	+28.6%	+97.6%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	+5.3%	+237.5%
Arm (ARM)	US Stock High Volatility	2023-11	+29.9%	+228.2%
Zhejiang Oriental (600120.SS)	A-Share Theme High Volatility	2025-01	-0.2%	+20.2%
MeiG Smart (002881.SZ)	A-Share Theme High Volatility	2025-01	-4.4%	+10.5%
UCloud (688158.SS)	A-Share Theme High Volatility	2025-01	+16.6%	+90.9%
QingCloud Technology (688316.SS)	A-Share Theme High Volatility	2025-01	+24.9%	+132.1%
Talkweb Information (002261.SZ)	A-Share Theme High Volatility	2025-01	+4.8%	+92.3%

Google brings Gemini CLI to developers as an open-source AI Agent tool. The terminal is becoming a direct entry point for model invocation, search, and code modification. Google: 2025 AI News Recap
OpenAI releases o3-pro. Reasoning models with higher compute budgets continue to place reliability, speed, and price on the same usage ledger. OpenAI Model Release Notes

Verdict of the month: Terminals and IDEs are no longer just the display interfaces for models, but the working environments where Agents obtain context, permissions, and acceptance feedback.

July: Agent Moves from Model to Delivery (688256.SS +110.4%)

Market Records: The AI rally has extended from chips to coding, software, and data center infrastructure. Leading U.S. stocks are rotating at high levels, while A-shares in optical modules, servers, and PCBs continue to diverge; high-beta targets such as Zhongji Innolight, Eoptolink, and Victory Giant Technology have become the “meme stock”-type representatives drawing market attention.

Monster Stocks Cumulative Table (as of the end of July 2025, month-end adjusted close composite; NVIDIA serves as the bellwether benchmark)

Target (Code)	Type	Inclusion Month	Monthly Change	Cumulative Since Inclusion
NVIDIA (NVDA)	Core Benchmark	2022-11	+12.6%	+1215.7%
C3.ai (AI)	US High-Volatility	2023-01	-4.1%	+111.2%
BigBear.ai (BBAI)	US High-Volatility	2023-01	-6.5%	+842.0%
Kunlun Tech (300418.SZ)	A-Share High-Volatility	2023-01	+17.0%	+156.2%
SoundHound AI (SOUN)	US High-Volatility	2023-02	-3.7%	+420.0%
Eoptolink Technology (300502.SZ)	A-Share High-Volatility	2023-02	+88.3%	+2090.1%
Cambricon (688256.SS)	A-Share High-Volatility	2023-02	+110.4%	+1676.9%
Foxconn Industrial Internet (601138.SS)	A-Share High-Volatility	2023-02	+55.5%	+501.6%
Innolight (300308.SZ)	A-Share High-Volatility	2023-03	+63.1%	+507.1%
Victory Giant Technology (300476.SZ)	A-Share High-Volatility	2023-04	+39.2%	+1375.2%
Super Micro Computer (SMCI)	US High-Volatility	2023-05	+20.3%	+457.8%
Palantir (PLTR)	US High-Volatility	2023-05	+16.2%	+1941.6%
Marvell (MRVL)	US High-Volatility	2023-05	+3.9%	+105.3%
Changshan Beiming (000158.SZ)	A-Share High-Volatility	2023-08	+11.5%	+276.3%
Arm (ARM)	US High-Volatility	2023-11	-12.6%	+186.8%
Zhejiang Orient (600120.SS)	A-Share Theme High-Volatility	2025-01	+6.6%	+28.2%
Meig Smart (002881.SZ)	A-Share Theme High-Volatility	2025-01	+27.8%	+41.2%
UCloud (688158.SS)	A-Share Theme High-Volatility	2025-01	+5.5%	+101.4%
QingCloud (688316.SS)	A-Share Theme High-Volatility	2025-01	-5.1%	+120.2%
Talkweb Information (002261.SZ)	A-Share Theme High-Volatility	2025-01	+27.7%	+145.6%

xAI released Grok 4. Reasoning, tool use, and real-time search are packaged into a single model product, extending frontier model competition further toward Agents and real-time information. xAI Newsroom
OpenAI released ChatGPT agent. Research, browser operation, terminal, and connectors are merged into a single Agent capable of executing complex tasks, with models beginning to directly handle requests from “start to delivery.” OpenAI: ChatGPT agent

Judgment for this month: The evaluation of Agents is beginning to approach that of software engineering, where consistently completing tasks matters more than producing a beautiful answer in one shot.

August: GPT-5 and Open Weights Advance Simultaneously (SOUN +26.0%)

Market Records: The U.S. stock market repriced model capabilities and platform ecosystems, with large-cap companies such as NVIDIA, Microsoft, and Meta outperforming most small-cap stocks; in the A-share market, AI capital spread from computing power to applications and end devices, without forming any single, sustainable “demon stock.”

Monster Stock Accumulation Table (as of August 2025, month-end adjusted closing compound; NVIDIA as benchmark leader)

Ticker (Code)	Type	Inclusion Month	Monthly Change	Cumulative Since Inclusion
NVIDIA (NVDA)	Core Bellwether Reference	2022-11	-2.1%	+1188.1%
C3.ai (AI)	US High-Volatility	2023-01	-28.2%	+51.7%
BigBear.ai (BBAI)	US High-Volatility	2023-01	-20.2%	+651.7%
Kunlun Tech (300418.SZ)	A-Share High-Volatility	2023-01	+15.7%	+196.4%
SoundHound AI (SOUN)	US High-Volatility	2023-02	+26.0%	+555.2%
Eoptolink (300502.SZ)	A-Share High-Volatility	2023-02	+2.7%	+2149.2%
Cambricon (688256.SS)	A-Share High-Volatility	2023-02	-11.2%	+1477.9%
Foxconn Industrial Internet (601138.SS)	A-Share High-Volatility	2023-02	+22.6%	+637.6%
Innolight (300308.SZ)	A-Share High-Volatility	2023-03	+13.7%	+590.3%
Victory Giant (300476.SZ)	A-Share High-Volatility	2023-04	+6.8%	+1475.5%
Super Micro (SMCI)	US High-Volatility	2023-05	-29.6%	+292.7%
Palantir (PLTR)	US High-Volatility	2023-05	-1.0%	+1921.2%
Marvell (MRVL)	US High-Volatility	2023-05	-21.8%	+60.5%
Changshan Beiming (000158.SZ)	A-Share High-Volatility	2023-08	-11.7%	+232.3%
Arm (ARM)	US High-Volatility	2023-11	-2.2%	+180.5%
Zhejiang Orient (600120.SS)	A-Share Thematic High-Volatility	2025-01	-2.5%	+25.0%
Meig Smart (002881.SZ)	A-Share Thematic High-Volatility	2025-01	-13.6%	+22.0%
UCloud (688158.SS)	A-Share Thematic High-Volatility	2025-01	-2.8%	+95.8%
QingCloud (688316.SS)	A-Share Thematic High-Volatility	2025-01	-11.5%	+94.9%
Talkweb Information (002261.SZ)	A-Share Thematic High-Volatility	2025-01	-12.8%	+114.2%

OpenAI releases GPT-5. A unified system brings fast answers, deep reasoning, vision, code, and Agent capabilities into a single generation of product, with coding explicitly positioned as a core use case. OpenAI: GPT-5
OpenAI releases gpt-oss-120b and gpt-oss-20b. OpenAI re-enters the open-weight model market, with models that can be downloaded, customized, and run locally or on one’s own infrastructure. OpenAI: gpt-oss

Judgment of the month: Closed-source services and open-weight models are not an either-or choice; model companies may charge via API while simultaneously expanding their ecosystem through open weights.

September: Agent Needs Checkpointing and Rollback (MRVL +33.7%)

Market Recap: Interest rate cuts and software earnings expectations have caused AI trading to shift from hardware to platforms, software, and data, with amplified volatility in Palantir, software infrastructure, and certain robotics stocks; A-share high-level thematic stocks have entered a phase of rapid rises and sharp declines.

Monster Stock Cumulative Table (as of 2025-09, month-end adjusted close composite; Nvidia as the benchmark leader)

Target (Code)	Type	Listed Month	Monthly Change	Cumulative Since Listing
NVIDIA (NVDA)	Core Benchmark	2022-11	+7.1%	+1279.6%
C3.ai (AI)	US Stock - High Volatility	2023-01	+2.5%	+55.5%
BigBear.ai (BBAI)	US Stock - High Volatility	2023-01	+28.6%	+866.7%
Kunlun Tech (300418.SZ)	A-Share - High Volatility	2023-01	-5.2%	+181.0%
SoundHound AI (SOUN)	US Stock - High Volatility	2023-02	+23.5%	+709.2%
Eoptolink Technology (300502.SZ)	A-Share - High Volatility	2023-02	-5.9%	+2016.5%
Cambricon (688256.SS)	A-Share - High Volatility	2023-02	+3.8%	+1537.8%
Foxconn Industrial Internet (601138.SS)	A-Share - High Volatility	2023-02	+9.1%	+704.7%
Innolight (300308.SZ)	A-Share - High Volatility	2023-03	+17.3%	+709.7%
Victory Giant Technology (300476.SZ)	A-Share - High Volatility	2023-04	+3.0%	+1522.7%
Super Micro Computer (SMCI)	US Stock - High Volatility	2023-05	+15.4%	+353.2%
Palantir (PLTR)	US Stock - High Volatility	2023-05	+16.4%	+2252.7%
Marvell (MRVL)	US Stock - High Volatility	2023-05	+33.7%	+114.6%
Changshan Beiming (000158.SZ)	A-Share - High Volatility	2023-08	+0.9%	+235.2%
Arm (ARM)	US Stock - High Volatility	2023-11	+2.3%	+187.0%
Zhejiang Orient (600120.SS)	A-Share - Thematic High Volatility	2025-01	+14.0%	+42.5%
MeiG Smart (002881.SZ)	A-Share - Thematic High Volatility	2025-01	-7.2%	+13.2%
UCloud (688158.SS)	A-Share - Thematic High Volatility	2025-01	-6.7%	+82.7%
QingCloud Technology (688316.SS)	A-Share - Thematic High Volatility	2025-01	-3.3%	+88.5%
Talkweb Information (002261.SZ)	A-Share - Thematic High Volatility	2025-01	-5.2%	+103.1%

Anthropic releases Claude Sonnet 4.5. Coding, computer use, Claude Code checkpoints, and the Agent SDK were rolled out in the same batch — rollback capability for long-running tasks is starting to become a product feature. Anthropic: Claude Sonnet 4.5

This month’s judgment: For Agents entering production, checkpoints, memory, permissions, and rollback are just as important as model capabilities.

October: Low-latency model offloading (ARM +20.0%)

Market Notes: The commercialization prospects of low-latency models support software and cloud service stocks, while chip stocks are consolidating at high levels; the AI sector in A-shares continues to rotate, with speculative stocks shifting from single-model concepts to crossover themes such as low-altitude economy, robotics, and edge devices.

Monster Stocks Cumulative Table (as of 2025-10, month-end adjusted closing price composite; NVIDIA as benchmark leader)

Asset (Code)	Type	Listed Month	Monthly Change	Cumulative Since Listing
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	+8.5%	+1396.8%
C3.ai (AI)	US Stock High Volatility	2023-01	+1.4%	+57.6%
BigBear.ai (BBAI)	US Stock High Volatility	2023-01	+6.1%	+925.7%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	-5.6%	+165.3%
SoundHound AI (SOUN)	US Stock High Volatility	2023-02	+9.6%	+786.8%
Eoptolink (300502.SZ)	A-Share High Volatility	2023-02	+1.0%	+2037.7%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	-3.2%	+1485.4%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	-15.7%	+578.3%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	+8.8%	+780.9%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	-7.9%	+1394.5%
Super Micro Computer (SMCI)	US Stock High Volatility	2023-05	+8.4%	+391.2%
Palantir (PLTR)	US Stock High Volatility	2023-05	+9.9%	+2485.6%
Marvell (MRVL)	US Stock High Volatility	2023-05	+11.6%	+139.5%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	-5.1%	+218.1%
Arm (ARM)	US Stock High Volatility	2023-11	+20.0%	+244.4%
Zhejiang Orient (600120.SS)	A-Share Theme High Volatility	2025-01	-12.7%	+24.4%
Meig Smart (002881.SZ)	A-Share Theme High Volatility	2025-01	-8.3%	+3.8%
UCloud (688158.SS)	A-Share Theme High Volatility	2025-01	-4.3%	+74.8%
QingCloud (688316.SS)	A-Share Theme High Volatility	2025-01	-0.8%	+87.0%
Talkweb (002261.SZ)	A-Share Theme High Volatility	2025-01	-7.6%	+87.6%

Anthropic releases Claude Haiku 4.5. A faster, cheaper model that brings the previous generation’s frontier capabilities down to high-frequency tasks, with model routing and cost control continuing to segment. Anthropic: Claude Haiku 4.5

This month’s take: The industry won’t be left with just one strongest model — low-latency and low-cost models determine whether a large volume of real-world calls can hold up.

November: Gemini 3 Connected Development Platform (300502.SZ +23.9%)

Market Notes: Leading U.S. AI stocks continue to be caught in a tug-of-war between capital expenditure and valuation, with NVIDIA, Broadcom, and cloud providers as the main themes, while software stocks diverge; A-shares see rotation among computing power, applications, and robotics themes, with no single stock able to represent the entire AI market.

Monster Stock Cumulative Table (as of 2025-11, month-end adjusted close compound; NVIDIA as the bellwether reference)

Target (Code)	Type	Inclusion Month	Monthly Change	Cumulative Since Inclusion
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	-12.6%	+1208.2%
C3.ai (AI)	US Stock High Volatility	2023-01	-17.8%	+29.6%
BigBear.ai (BBAI)	US Stock High Volatility	2023-01	-8.4%	+839.5%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	-4.1%	+154.4%
SoundHound AI (SOUN)	US Stock High Volatility	2023-02	-31.6%	+506.6%
Eoptolink Technology (300502.SZ)	A-Share High Volatility	2023-02	+23.9%	+2548.6%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	+1.8%	+1514.0%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	+2.2%	+593.3%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	+18.6%	+944.8%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	+6.2%	+1487.2%
Super Micro Computer (SMCI)	US Stock High Volatility	2023-05	-34.9%	+219.8%
Palantir (PLTR)	US Stock High Volatility	2023-05	-16.0%	+2071.9%
Marvell (MRVL)	US Stock High Volatility	2023-05	-4.6%	+128.5%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	-6.8%	+196.5%
Arm (ARM)	US Stock High Volatility	2023-11	-20.2%	+174.8%
Zhejiang Orient (600120.SS)	A-Share Thematic High Volatility	2025-01	-1.3%	+22.7%
MeiG Smart (002881.SZ)	A-Share Thematic High Volatility	2025-01	+6.9%	+11.0%
UCloud (688158.SS)	A-Share Thematic High Volatility	2025-01	+16.3%	+103.3%
QingCloud Technology (688316.SS)	A-Share Thematic High Volatility	2025-01	-5.6%	+76.5%
Talkweb Information (002261.SZ)	A-Share Thematic High Volatility	2025-01	+4.5%	+96.1%

Google releases Gemini 3. Reasoning, multimodal, and coding are reintegrated, and the agentic development platform begins to become part of product distribution. Google: Gemini 3

Judgment for This Month: The battleground for model vendors has expanded from model APIs to development platforms, search, browsers, and office software.

December: Inference Capabilities Continue to Drive Down Costs (688158.SS +33.6%)

Market Notes: At year-end, the market places greater emphasis on inference costs and real-world usage. AI leaders are shifting from a pure hardware-driven narrative to a layered structure of cloud, software, and applications. High-volatility speculative stocks still exist, but their duration is shortening, and they should not be used as proxies for industry prosperity.

Monster Stocks Cumulative Table (As of 2025-12, Month-end Adjusted Close Composite; NVIDIA as Leading Benchmark)

Ticker (Code)	Type	Listed Month	Monthly Change	Cumulative Since Listing
NVIDIA (NVDA)	Core Benchmark	2022-11	+5.4%	+1278.9%
C3.ai (AI)	US Stock, High Volatility	2023-01	-6.7%	+20.9%
BigBear.ai (BBAI)	US Stock, High Volatility	2023-01	-14.8%	+700.5%
Kunlun Tech (300418.SZ)	A-Share, High Volatility	2023-01	+33.8%	+240.4%
SoundHound AI (SOUN)	US Stock, High Volatility	2023-02	-17.3%	+401.7%
Eoptolink (300502.SZ)	A-Share, High Volatility	2023-02	-2.6%	+2479.7%
Cambricon (688256.SS)	A-Share, High Volatility	2023-02	-7.1%	+1399.4%
Foxconn Industrial Internet (601138.SS)	A-Share, High Volatility	2023-02	-6.5%	+548.2%
Innolight (300308.SZ)	A-Share, High Volatility	2023-03	+6.4%	+1011.7%
Victory Giant Technology (300476.SZ)	A-Share, High Volatility	2023-04	-7.9%	+1361.8%
Super Micro Computer (SMCI)	US Stock, High Volatility	2023-05	-13.5%	+176.6%
Palantir (PLTR)	US Stock, High Volatility	2023-05	+5.5%	+2191.4%
Marvell (MRVL)	US Stock, High Volatility	2023-05	-4.9%	+117.3%
Changshan Beiming (000158.SZ)	A-Share, High Volatility	2023-08	-1.7%	+191.5%
Arm (ARM)	US Stock, High Volatility	2023-11	-19.4%	+121.5%
Zhejiang Orient (600120.SS)	A-Share, Thematic High Volatility	2025-01	+5.4%	+29.4%
MeiG Smart (002881.SZ)	A-Share, Thematic High Volatility	2025-01	+0.4%	+11.4%
UCloud (688158.SS)	A-Share, Thematic High Volatility	2025-01	+33.6%	+171.7%
QingCloud (688316.SS)	A-Share, Thematic High Volatility	2025-01	+25.9%	+122.2%
Talkweb Information (002261.SZ)	A-Share, Thematic High Volatility	2025-01	-4.3%	+87.6%

Google releases Gemini 3 Flash. A faster and more cost-effective model carries forward the capabilities of Gemini 3, as the price gap between frontier models and large-scale deployments continues to narrow. Google: December 2025 AI Update

This Month’s Takeaway: Reasoning capabilities will only transition from demos to everyday workflows once the per-unit cost is driven down enough.

2026: Model companies begin to face scrutiny from the public market and real-world workflows

January: Constitution and Automatic Browsing (NVDA +2.5%)

Market Records: AI companies in the Hong Kong stock market now have publicly tradable samples, and Zhipu (02513) and MiniMax (00100) have shown significant volatility after listing; the U.S. stock market continues to be dominated by Nvidia, Microsoft, and cloud providers. The high turnover of newly listed stocks can be recorded as “zombie stock”-style market action, but it cannot replace revenue and valuation analysis. Hong Kong Stock Exchange: MiniMax Newly Listed Securities Admission Notice

Monster Stocks Cumulative Table (as of 2026-01, month-end adjusted closing composite; Nvidia as the bellwether for comparison)

Target (Code)	Type	Listed Month	Monthly Change	Cumulative Since Listing
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	+2.5%	+1313.4%
C3.ai (AI)	US High-Volatility	2023-01	-18.3%	-1.2%
BigBear.ai (BBAI)	US High-Volatility	2023-01	-6.7%	+646.9%
Kunlun Tech (300418.SZ)	A-Share High-Volatility	2023-01	+9.5%	+272.7%
SoundHound AI (SOUN)	US High-Volatility	2023-02	-15.1%	+325.9%
Eoptolink (300502.SZ)	A-Share High-Volatility	2023-02	-14.2%	+2113.4%
Cambricon (688256.SS)	A-Share High-Volatility	2023-02	-6.4%	+1303.4%
Foxconn Industrial Internet (601138.SS)	A-Share High-Volatility	2023-02	-3.5%	+525.5%
Innolight (300308.SZ)	A-Share High-Volatility	2023-03	-17.7%	+814.9%
Victory Giant Technology (300476.SZ)	A-Share High-Volatility	2023-04	+14.8%	+1578.2%
Super Micro Computer (SMCI)	US High-Volatility	2023-05	-0.5%	+175.2%
Palantir (PLTR)	US High-Volatility	2023-05	-17.5%	+1790.4%
Marvell (MRVL)	US High-Volatility	2023-05	-7.1%	+101.9%
Changshan Beiming (000158.SZ)	A-Share High-Volatility	2023-08	+4.9%	+205.8%
Arm (ARM)	US High-Volatility	2023-11	-3.6%	+113.5%
Zhejiang Orient (600120.SS)	A-Share Theme High-Volatility	2025-01	+2.9%	+33.1%
Meig Smart (002881.SZ)	A-Share Theme High-Volatility	2025-01	+9.9%	+22.5%
UCloud (688158.SS)	A-Share Theme High-Volatility	2025-01	+7.6%	+192.3%
QingCloud (688316.SS)	A-Share Theme High-Volatility	2025-01	+9.7%	+143.8%
Talkweb Information (002261.SZ)	A-Share Theme High-Volatility	2025-01	+15.3%	+116.3%

Anthropic publishes Claude’s new Constitution. Model values, training goals, and behavioral boundaries are now publicly available as a fully readable document, with safety and governance moving from the appendix into product materials. Anthropic: Claude’s Constitution
Google summarizes its January 2026 AI updates. Gemini further integrates into Chrome, Search, and automated browsing, with AI products evolving from answering questions to handling multi-step tasks. Google: AI Updates January 2026

Judgment of the Month: The competitive focus of 2026 has shifted to “whether you can advance a task for the user” rather than “whether you can generate another paragraph.”

February: Reasoning Models Move Toward Research Collaboration (2513.HK +20.6%)

Market Records: Hong Kong stock trading of Zhipu and MiniMax has shifted from listing sentiment to circulating shares, unlocking, and valuation games, with a divergence in Hong Kong tech stocks; U.S. AI leaders are still priced around capital expenditure and inference demand, with no new globally unified “monster stock” emerging for now. Hong Kong Exchange: Zhipu and MiniMax Approved for Short Selling

Monster Stocks Cumulative Table (as of 2026-02, month-end adjusted close composite; NVIDIA as the bellwether reference)

Asset (Code)	Type	Listing Month	Monthly Change	Cumulative Since Listing
Nvidia (NVDA)	Core Bellwether	2022-11	-7.3%	+1210.2%
C3.ai (AI)	US High-Volatility	2023-01	-27.8%	-28.7%
BigBear.ai (BBAI)	US High-Volatility	2023-01	-21.4%	+487.0%
Kunlun Tech (300418.SZ)	A-Share High-Volatility	2023-01	-19.8%	+198.9%
SoundHound AI (SOUN)	US High-Volatility	2023-02	+1.7%	+333.1%
Eoptolink (300502.SZ)	A-Share High-Volatility	2023-02	+23.1%	+2624.7%
Cambricon (688256.SS)	A-Share High-Volatility	2023-02	-16.6%	+1070.4%
Foxconn Industrial Internet (601138.SS)	A-Share High-Volatility	2023-02	-7.6%	+478.0%
Innolight (300308.SZ)	A-Share High-Volatility	2023-03	+6.6%	+875.3%
Victory Giant Technology (300476.SZ)	A-Share High-Volatility	2023-04	-17.5%	+1284.5%
Super Micro Computer (SMCI)	US High-Volatility	2023-05	+11.3%	+206.3%
Palantir (PLTR)	US High-Volatility	2023-05	-6.4%	+1669.4%
Marvell (MRVL)	US High-Volatility	2023-05	+3.5%	+108.9%
Changshan Beiming (000158.SZ)	A-Share High-Volatility	2023-08	-17.2%	+153.2%
Arm (ARM)	US High-Volatility	2023-11	+21.0%	+158.3%
Zhejiang Orient (600120.SS)	A-Share Thematic High-Volatility	2025-01	-16.1%	+11.7%
Meig Smart (002881.SZ)	A-Share Thematic High-Volatility	2025-01	-22.2%	-4.7%
UCloud (688158.SS)	A-Share Thematic High-Volatility	2025-01	-0.4%	+191.1%
QingCloud Technology (688316.SS)	A-Share Thematic High-Volatility	2025-01	-16.7%	+103.1%
Talkweb Information (002261.SZ)	A-Share Thematic High-Volatility	2025-01	-6.4%	+102.5%
Zhipu (2513.HK)	HK-Share IPO High-Volatility	2026-02	+20.6%	+20.6%

Anthropic releases Claude Opus 4.6. Coding, Agents, computer use, tool calling, and search are pushed into higher capability tiers, as model companies continue to compete for complex workflows. Anthropic Newsroom
Google updates Gemini 3 Deep Think. The dedicated deep-reasoning mode is extended toward science, research, and engineering tasks. Google: Gemini 3 Deep Think

Judgment for this month: Reasoning models are shifting from exams and leaderboards toward research collaboration, but the real value still depends on whether experts can review and reuse them.

March: Agent Completes Retrieval and Memory (688256.SS +72.9%)

Market Record: The divergence among AI stocks continues to widen. Large-cap platforms and computing power suppliers in the U.S. stock market remain relatively stable, while the intraday volatility of Zhipu and MiniMax in the Hong Kong stock market, as well as the computing power chain in the A-share market, is greater; “demon stocks” have shifted from single model names to targets with smaller circulating shares and rapidly changing expectations.

Meme Stock Cumulative Table (As of 2026-03, month-end adjusted close composite; NVIDIA as the benchmark leader)

Target (Code)	Type	Listed Month	Monthly Change	Cumulative Since Listing
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	-1.6%	+1189.2%
C3.ai (AI)	US Stock High Volatility	2023-01	+5.9%	-24.5%
BigBear.ai (BBAI)	US Stock High Volatility	2023-01	-11.1%	+421.9%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	-0.3%	+198.0%
SoundHound AI (SOUN)	US Stock High Volatility	2023-02	-20.1%	+246.1%
Eoptolink (300502.SZ)	A-Share High Volatility	2023-02	+18.7%	+3134.2%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	+72.9%	+1923.7%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	+22.2%	+606.3%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	+50.8%	+1370.7%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	+31.7%	+1723.4%
Super Micro Computer (SMCI)	US Stock High Volatility	2023-05	-29.7%	+115.4%
Palantir (PLTR)	US Stock High Volatility	2023-05	+6.6%	+1786.2%
Marvell (MRVL)	US Stock High Volatility	2023-05	+21.3%	+153.4%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	-1.1%	+150.4%
Arm (ARM)	US Stock High Volatility	2023-11	+18.7%	+206.7%
Zhejiang Orient (600120.SS)	A-Share Theme High Volatility	2025-01	+0.0%	+11.7%
Meig Smart (002881.SZ)	A-Share Theme High Volatility	2025-01	+2.2%	-2.6%
UCloud (688158.SS)	A-Share Theme High Volatility	2025-01	+12.0%	+226.1%
QingCloud Technology (688316.SS)	A-Share Theme High Volatility	2025-01	+0.9%	+104.9%
Talkweb Information (002261.SZ)	A-Share Theme High Volatility	2025-01	+3.7%	+110.0%
Zhipu (2513.HK)	HK Stock New Listing High Volatility	2026-02	+25.2%	+51.0%

Google released Gemini Embedding 2. Text, images, video, audio, and documents enter a unified multimodal vector space, with retrieval and classification beginning to become fundamental layer capabilities of Agents. Google: Gemini Embedding 2
Gemini’s Search Live, Canvas, and cross-application capabilities continue to expand. Search begins to simultaneously handle real-time conversation, long-task workspaces, and personal profile connections. Google: AI Updates March 2026

This month’s takeaway: The infrastructure for an Agent is not just a set of large models, but also includes retrieval, memory, file system, search, and cross-application permissions.

April: Long Tasks Encounter Product Life Cycle (2513.HK +83.8%)

Market Notes: The market has begun incorporating the retention, cost, and cash flow of AI products into stock price narratives, and the valuation elasticity of concept stocks has decreased. Leading U.S. stocks, Hong Kong-listed AI IPOs, and A-share computing power stocks are each following their own trends, without forming new cross-market monster stocks.

Monster Stock Cumulative Table (as of 2026-04, month-end adjusted closing composite; NVIDIA as the leading benchmark)

Asset (Code)	Type	Added	This Month	Cumulative Since Inclusion
NVIDIA (NVDA)	Core Bellwether	2022-11	+14.4%	+1374.9%
C3.ai (AI)	US High Volatility	2023-01	+4.9%	-20.8%
BigBear.ai (BBAI)	US High Volatility	2023-01	+13.1%	+490.2%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	-10.1%	+167.9%
SoundHound AI (SOUN)	US High Volatility	2023-02	+15.9%	+301.1%
Eoptolink (300502.SZ)	A-Share High Volatility	2023-02	+34.4%	+4246.7%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	+15.0%	+2227.3%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	+16.7%	+724.2%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	+35.4%	+1891.4%
Victory Giant (300476.SZ)	A-Share High Volatility	2023-04	+12.2%	+1945.8%
Super Micro Computer (SMCI)	US High Volatility	2023-05	+20.3%	+159.1%
Palantir (PLTR)	US High Volatility	2023-05	-4.9%	+1693.7%
Marvell (MRVL)	US High Volatility	2023-05	+66.8%	+322.7%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	-10.9%	+123.1%
Arm (ARM)	US High Volatility	2023-11	+39.0%	+326.3%
Zhejiang Oriental (600120.SS)	A-Share Theme High Volatility	2025-01	-13.8%	-3.7%
MeiG Smart (002881.SZ)	A-Share Theme High Volatility	2025-01	+16.3%	+13.3%
UCloud (688158.SS)	A-Share Theme High Volatility	2025-01	-14.1%	+180.1%
QingCloud (688316.SS)	A-Share Theme High Volatility	2025-01	-11.5%	+81.3%
Tarcogent Information (002261.SZ)	A-Share Theme High Volatility	2025-01	-16.3%	+75.8%
Zhipu (2513.HK)	HK IPO High Volatility	2026-02	+83.8%	+177.5%

Anthropic released Claude Opus 4.7. Stronger coding, Agent, vision, and multi-step task capabilities continue to push model competition toward long-horizon tasks. Anthropic Newsroom
Sora product discontinued. The technical impact of generative video persists, but the user base, cost, and commercial return of a single product are not automatically guaranteed by model demos. OpenAI: Service status note on the Sora release page

Judgment of the Month: AI products must also undergo scrutiny based on lifecycle and unit economics; having strong model capabilities does not guarantee that the product will endure in the long term.

May: AI Applications Become Execution Systems (SMCI +68.2%)

Market Recap: AI capital expenditure remains the main theme in U.S. stocks, but capital is starting to seek out software and applications that can convert calls into revenue; in A-shares, computing power, optical modules, and PCBs are rotating, while Hong Kong AI IPOs are highly volatile, and the gap between meme stocks and leading stocks is becoming more pronounced.

Monster Stocks Cumulative Table (as of 2026-05, month-end adjusted close composite; NVIDIA as benchmark leader)

Target (Code)	Type	Inclusion Month	Monthly Change	Cumulative Since Inclusion
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	+5.8%	+1460.4%
C3.ai (AI)	US Stock High Volatility	2023-01	+22.0%	-3.3%
BigBear.ai (BBAI)	US Stock High Volatility	2023-01	+26.6%	+647.2%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	-8.1%	+146.2%
SoundHound AI (SOUN)	US Stock High Volatility	2023-02	+13.1%	+353.7%
Eoptolink (300502.SZ)	A-Share High Volatility	2023-02	+20.5%	+5137.8%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	+21.8%	+2734.6%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	-1.7%	+710.2%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	+9.4%	+2078.5%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	-6.0%	+1823.1%
Super Micro Computer (SMCI)	US Stock High Volatility	2023-05	+68.2%	+335.8%
Palantir (PLTR)	US Stock High Volatility	2023-05	+12.5%	+1918.0%
Marvell (MRVL)	US Stock High Volatility	2023-05	+24.1%	+424.6%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	-15.7%	+88.1%
Arm (ARM)	US Stock High Volatility	2023-11	+68.0%	+616.1%
Zhejiang Orient (600120.SS)	A-Share Thematic High Volatility	2025-01	-9.9%	-13.2%
Meig Smart (002881.SZ)	A-Share Thematic High Volatility	2025-01	-19.8%	-9.2%
UCloud (688158.SS)	A-Share Thematic High Volatility	2025-01	-5.0%	+166.1%
QingCloud Technology (688316.SS)	A-Share Thematic High Volatility	2025-01	-12.7%	+58.3%
Talkweb Information (002261.SZ)	A-Share Thematic High Volatility	2025-01	-5.6%	+65.9%
Zhipu (2513.HK)	HK Stock IPO High Volatility	2026-02	+31.9%	+266.1%

Google I/O 2026 releases Gemini 3.5 Flash and Gemini Omni. Gemini enters the “agentic era,” with models simultaneously handling input, reasoning, action, and video creation. Google: I/O 2026 Overview
Google integrates Agent, Search, Gemini App, and developer platform into the same product update cycle. The boundaries between search, shopping, code, and personal assistant continue to blur. Google: I/O 2026

Judgment of the month: “AI applications” are increasingly looking less like a standalone chat box and more like an execution system that overlays search, office work, devices, and development tools.

June: On-device Models Meet the Open Market (MRVL +45.3%)

Market Recap: On-device models have brought Apple, mobile phone, and storage supply chains back into AI pricing, while the public market has begun scrutinizing model companies’ financing and losses. A-shares such as Zhongji Innolight, Eoptolink, and Victory Giant Technology have experienced high-level volatility, and the unlocks and changes in floating shares of Hong Kong-listed Zhipu and MiniMax have become new trading variables.

Monster Stock Cumulative Table (as of 2026-06, month-end adjusted closing composite; NVIDIA serves as the leader benchmark)

Target (Code)	Type	Listed Month	This Month Change	Cumulative Since Listing
NVIDIA (NVDA)	Core Leader Benchmark	2022-11	-5.1%	+1380.8%
C3.ai (AI)	US Stock High Volatility	2023-01	-15.6%	-18.4%
BigBear.ai (BBAI)	US Stock High Volatility	2023-01	-27.2%	+444.0%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	+22.4%	+201.4%
SoundHound AI (SOUN)	US Stock High Volatility	2023-02	-28.1%	+226.2%
Eoptolink (300502.SZ)	A-Share High Volatility	2023-02	-13.8%	+4415.0%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	-12.3%	+2385.9%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	-8.1%	+644.6%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	-13.9%	+1775.7%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	-21.5%	+1409.6%
Super Micro Computer (SMCI)	US Stock High Volatility	2023-05	-36.4%	+177.1%
Palantir (PLTR)	US Stock High Volatility	2023-05	-25.5%	+1403.4%
Marvell (MRVL)	US Stock High Volatility	2023-05	+45.3%	+662.3%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	+3.7%	+95.0%
Arm (ARM)	US Stock High Volatility	2023-11	+0.4%	+619.0%
Zhejiang Orient (600120.SS)	A-Share Theme High Volatility	2025-01	+12.6%	-2.3%
MeiG Smart (002881.SZ)	A-Share Theme High Volatility	2025-01	+3.5%	-6.0%
UCloud (688158.SS)	A-Share Theme High Volatility	2025-01	+1.2%	+169.3%
QingCloud Technology (688316.SS)	A-Share Theme High Volatility	2025-01	+0.9%	+59.7%
Talkweb Information (002261.SZ)	A-Share Theme High Volatility	2025-01	+14.5%	+90.0%
Zhipu (2513.HK)	HK-Share IPO High Volatility	2026-02	-22.1%	+185.2%

Apple Released the Third Generation of Apple Foundation Models. The on-device model, server model, and the foundation model system developed in collaboration with Google are incorporated into the new generation of Apple Intelligence. Apple Machine Learning Research
Anthropic Released Claude Sonnet 5. Lower-cost models begin to take on browser, terminal, tool-calling, and long-task execution, with Agent capabilities continuing to penetrate into larger-scale everyday usage. Anthropic: Claude Sonnet 5
Anthropic Submitted Confidential IPO Filing. The company has not yet determined the offering size, pricing, or timing of the listing; what is confirmed is that model companies are beginning to consider the public market as a viable capital path. AP: Anthropic confidential IPO filing

Judgment for this month: On-device models and the public market represent two different kinds of constraints: on one side, power consumption, privacy, and latency; on the other side, revenue, losses, financing, and cash flow.

July: Coding Agent Becomes a Clear Main Theme (NVDA +5.4%)

Market Notes: As of July 13th, leading US AI stocks continue to rotate at high levels around computing power, cloud, and software platforms. In the A-share market, optical modules and PCBs show greater momentum, while Hong Kong stocks Zhipu and MiniMax are even more volatile. Coding agents represent a new market narrative, but it has yet to be proven that they can drive all AI stocks higher in unison.

Monster Stock Cumulative Table (as of 2026-07, month-end adjusted closing composite; NVIDIA as benchmark leader)

Asset (Code)	Type	Inclusion Month	Monthly Change	Cumulative Since Inclusion
NVIDIA (NVDA)	Core Benchmark	2022-11	+5.4%	+1460.8%
C3.ai (AI)	US High Volatility	2023-01	-1.5%	-19.6%
BigBear.ai (BBAI)	US High Volatility	2023-01	-10.9%	+384.7%
Kunlun Tech (300418.SZ)	A-Share High Volatility	2023-01	-2.5%	+193.9%
SoundHound AI (SOUN)	US High Volatility	2023-02	+2.6%	+234.7%
Eoptolink (300502.SZ)	A-Share High Volatility	2023-02	-2.0%	+4324.7%
Cambricon (688256.SS)	A-Share High Volatility	2023-02	-0.7%	+2368.5%
Foxconn Industrial Internet (601138.SS)	A-Share High Volatility	2023-02	-4.8%	+608.9%
Innolight (300308.SZ)	A-Share High Volatility	2023-03	+1.3%	+1800.1%
Victory Giant Technology (300476.SZ)	A-Share High Volatility	2023-04	+0.4%	+1415.6%
Super Micro Computer (SMCI)	US High Volatility	2023-05	-3.5%	+167.4%
Palantir (PLTR)	US High Volatility	2023-05	+8.7%	+1534.2%
Marvell (MRVL)	US High Volatility	2023-05	-20.8%	+503.7%
Changshan Beiming (000158.SZ)	A-Share High Volatility	2023-08	-2.5%	+90.1%
Arm (ARM)	US High Volatility	2023-11	-8.8%	+555.7%
Zhejiang Orient (600120.SS)	A-Share Theme High Volatility	2025-01	+1.4%	-1.0%
MeiG Smart (002881.SZ)	A-Share Theme High Volatility	2025-01	-0.2%	-6.2%
UCloud (688158.SS)	A-Share Theme High Volatility	2025-01	-4.1%	+158.2%
QingCloud (688316.SS)	A-Share Theme High Volatility	2025-01	-6.2%	+49.8%
Talkweb Information (002261.SZ)	A-Share Theme High Volatility	2025-01	+2.9%	+95.5%
Zhipu (2513.HK)	HK IPO High Volatility	2026-02	+0.3%	+186.0%

July 8: xAI releases Grok 4.5. The launch page lists coding, agents, and knowledge work as the primary scenarios, with availability in Cursor, Grok Build, and the API. xAI: Grok 4.5
Anthropic releases a retrospective on Claude Code. The journey from an internal CLI to a coding agent is documented in public, with software engineering continuing to serve as an important workflow for validating long-horizon agents. Anthropic: The Making of Claude Code
OpenAI releases ChatGPT Work. Agents begin to carry complex projects forward across applications and files, as the product narrative shifts further from “answering questions” toward “breaking goals down into outcomes.” OpenAI: ChatGPT Work

Judgment for this month: As of July 13, the clearest product direction is still the coding Agent; its advantage lies not in being able to write code, but in the fact that tasks can be split, run, tested, and rolled back.

What to Watch After the Timeline

The Entry Point Has Shifted from Chat to Workflow

ChatGPT first proved that natural language could become the entry point for software. GPT-4, plugins, multimodal capabilities, and search then connected that entry point to files, the web, and external tools. After 2025, Deep Research, Claude Code, Codex, and various terminal Agents have continued to extend the task chain. What users now deliver is no longer just an answer, but a piece of research, a set of modifications, a collection of test results, or a project that can keep running.

Open Weights and Closed-Source Services Each Bear Different Costs

Route	What You Get	What It Costs	Suitable Problems
Closed-Source API	Fast updates, complete products, enterprise support, and service stability	Call fees, quotas, vendor dependency, and data boundaries	Need to launch quickly, handle complex capabilities, and deliver uniformly
Open Weights	Local deployment, fine-tuning, more controllable data boundaries	GPUs, memory, operations, licenses, and upgrade costs	Private data, local experimentation, and custom deployment
Hybrid Routing	Simple tasks handled locally, complex tasks routed to the cloud	Requires evaluation, routing, fallback, and monitoring	Workflows that care about cost, privacy, and capability simultaneously

“Open” does not mean free, and “closed-source” does not mean no ecosystem. Real deployment decisions come down to licensing, GPU memory, context length, inference speed, tool calling, fault handling, and accountability.

AIGC Moves from Content Generation to Credibility Competition

Text requires fact-checking; images require verification of source and copyright; videos require confirmation of timing and source material; and voice/audio requires consideration of identity boundaries. As the cost of generation decreases, the cost of verification becomes even more important. Visual completeness cannot replace the chain of evidence, and the model’s confident tone cannot replace original documents and reproducible processes.

Coding is the earliest Agent scenario that can be accepted

Code completion is only the starting point. A real coding Agent needs to read repositories, understand constraints, modify files, execute commands, run tests, handle failures, retain checkpoints, and allow humans to review and roll back. Software engineering has clear delivery boundaries, so it’s easier to form subscriptions or enterprise budgets than general chat; but this also means every line of code generated by the model must be tested and maintained.

AI Stocks Need to Move from Computing Power Orders to Cash Flow

AI stocks do not all rise in a straight line together. In the early market, investors bought the entry point and the vision; starting in 2023, they bought GPU orders and data centers; and from 2025 onward, they care more about unit inference cost, application-level payments, capital expenditure, and the financing ability of model companies. The upstream can receive money first, while the application layer still has to go through pilots, procurement, renewals, and profit validation.

Looking back at the monthly “Market Records,” three clear re-anchorings can be observed: before May 2023, the market was pricing in “whether generative AI would become the gateway”; after NVIDIA’s earnings report, compute orders and data centers became the primary theme; after the DeepSeek shock, the market began to ask again whether training and inference could become cheaper. By 2026, model companies such as Zhipu and MiniMax entered the Hong Kong stock market, and the question shifted to revenue, losses, free float, and lock-up expiry, rather than simply comparing model launch events.

The following three tables are worth tracking more closely:

The capital expenditure, orders, and remaining contract liabilities of cloud vendors determine how long infrastructure demand can keep rolling.
The revenue, losses, inference costs, customer concentration, and cash flow of model companies determine whether capability can be turned into budget.
The usage depth, rework rate, test pass rate, and renewal rate of coding and Agent products determine whether “being able to complete a task” truly equals “having commercial value”.

Market Data and Record Boundaries

The monthly commentary and cumulative tables in this article are based on Yahoo Finance’s publicly available monthly data feeds and news from the corresponding months. The tables use adjusted closing prices. For the current month, data is current through July 13, and indices, currencies, and trading regimes from different markets have not been forcibly merged. Since Yahoo Finance does not provide an available monthly series for MiniMax (00100.HK), this article does not fabricate price changes for it. The term “妖股” (meme stock) is used here only to look back on market sentiment and does not constitute investment advice. For matters involving listing dates, short-selling eligibility, share unlock schedules, and corporate disclosures, please refer to the official announcements from the relevant exchange.

References

This article only compiles publicly available materials and personal observations, and does not constitute investment advice. Models, products, prices, stock prices, and company disclosures are subject to continuous change; for content related to July 2026, the cut-off date is July 13th, and you should refer to the company’s latest announcements and original market data thereafter.

写作附记

Original Prompt

The current AI Big Events are mainly an aggregation of my historical articles, lacking a corresponding timeline. For each month, I need to record AI-related news, perform online searches, and add a wave of monthly hot events, controlled within 3 per month. The article structure should also be adjusted accordingly.

Original Prompt of the First Draft

Refer to the major events of 2026, create new AI major events starting from the emergence of ChatGPT, delete the content from the major events of 2026 and 2025 that you think should be removed in the 2025 draft, and extract them into AI major events. Search online for what content should be included in AI major events: the development of large model capabilities, open-source models, closed-source models, stock price changes of related companies in the stock market, AIGC has now started competing in coding, there is a lot of content, please organize the outline clearly.

Raw Prompt for This Round

AI major events are now recorded monthly as key events, but the corresponding stock market movements for these events are not recorded. During this wave of AI, the stock market has also been historically rare. A new event is added monthly at the top, simply recording that month’s stock trends and noting various “monster stocks.”

This round keeps the topic classification of the original manuscript after the timeline, and restructures the main sections to be organized chronologically by year and month. The monthly events are primarily based on publicly released pages, and do not present predictions, rumors, or proposed financing as established facts.

Trump Visits China Again: This China-US Leaders Meeting, First Seeking Stability, Then Discussing Transactions

Thu, 14 May 2026 10:09:01 +0800

This meeting is neither about the sudden “thaw” of Sino-US relations, nor is it about one side completely overpowering the other. Rather, it appears to be a mandatory recalibration conducted under heightened pressure.

If you only look at the surface, you see welcome ceremonies, state banquets, the Temple of Heaven, and corporate delegations—it is very spectacular. But if you lay out the timeline, what this meeting truly needs to address are four tougher matters: how to maintain the trade and economic ceasefire left over from the Busan talks last October; how to manage the risks concerning Taiwan; how to mitigate losses stemming from the spillovers of the Iran conflict; and whether, under their respective domestic political pressures, both sides can first stabilize relations and avoid continued decline.

In other words, this Beijing meeting first and foremost aimed at “preventing a loss of control,” with “making deals” being secondary.

To clarify the scope: This article compiles information based on what has been publicly released up to May 14, 2026 morning. Since this visit is still ongoing, the content covers the background details, already announced itineraries, and external expectations, rather than presenting unhappened results as established facts.

Let’s first clarify this meeting.

This meeting corresponded to a state visit to China by US President Donald Trump from May 13 to May 15, 2026. According to the Ministry of Foreign Affairs of China’s statement, this was “the first US presidential visit to China in nearly nine years,” and it was also Trump’s second visit as president since November 2017.

According to public information, this visit has at least three layers:

Level	Key Discussions	Implication
Head of State Level	Xi Jinping and Trump meeting face-to-face in Beijing.	High-level guidance sets the tone; priority is preventing bilateral relations from continuing to drift.
Economic/Trade Level	He Lifeng and Besant holding pre-consultations in Korea.	Indicates that the discussions concern not merely protocol, but specific transactions and exchange terms.
External/Global Level	International attention highly focused on Iran, Taiwan, technology restrictions, agricultural products, and Boeing orders.	The importance of the meeting has long exceeded bilateral issues, carrying global market and geopolitical spillover effects.

So, this isn’t just an ordinary courtesy visit. It’s a high-level meeting dealing with specific, weighty agenda items, while still needing to maintain a dignified atmosphere.

The whole story was not actually sudden

If you only look at Trump arriving in Beijing on May 13th, it might seem like this meeting came very quickly. In reality, several rounds of groundwork had already been laid beforehand.

First Paragraph Setup: The Busan meeting in October 2025 needs to slow things down.

The Chinese Ministry of Foreign Affairs and the Ministry of Commerce both repeatedly mentioned one phrase: “Busan Talks”.

This is very crucial. This is because many of Beijing’s public statements this time treat “implementing the important consensus from the Busan meeting and past talks” as a prerequisite. Foreign media also mentioned that after the Busan meeting in October last year, both sides had established a round of economic and trade ceasefire arrangements: the US suspended the three-digit level tariff escalation on Chinese goods, while China did not restrict its rare earth supply to a more severe extent.

My judgment is that the Busan talks did not solve the problem, but rather deferred the most dangerous escalation button for now. This meeting in Beijing was fundamentally about confirming whether this temporary ceasefire mechanism could be sustained.

Preamble for the Second Section: Leaders’ Call on `February 4, 2026`, putting this year’s agenda forward first

On February 4, 2026, Xi Jinping had a phone call with Trump. There are two noteworthy points in the official Chinese statement.

First, it explicitly mentions that communication between both sides was smooth last year, and the Busan meeting “pointed the direction and course for US-China relations.” This demonstrates that Beijing’s narrative is consistent: emphasizing the leadership guiding the way first, rather than focusing on specific transactions first.

Secondly, the call directly highlighted each side’s important agendas in 2026, including China initiating the “15th Five-Year Plan,” China hosting APEC, and the US hosting the G20. The subtext of this statement is clear: neither country can afford to push the relationship towards an uncontrollable state this year.

Third Section Preparation: The ROK Economic and Trade Consultation on May 12-13 Provides Background for the Beijing Talks

What genuinely shows is that this meeting is not only for symbolic discussion, but also serves as pre-meeting consultations held in Korea.

On May 10th, China’s Ministry of Commerce announced that He Lifeng would travel to South Korea on May 12th–13th to hold economic and trade consultations with the U.S. side, explicitly stating that this was “guided by the important consensus reached during the summit meeting and previous calls between the two heads of state.” By May 13th, Xinhua News Agency provided another very standard but information-rich summary: the two sides conducted “candid, in-depth, and constructive exchanges.”

Such wording typically indicates that while negotiations have not reached an impasse, they haven’t achieved enough major breakthroughs to be announced prematurely. It is more like preemptively clearing away the technical hurdles that need to be addressed before a high-level summit.

Contextual Background: The very timing of the visit to China itself also suggests that the situation is more pressing than what appears on the surface

Foreign media disclosed that this visit to China was originally supposed to take place earlier, but was subsequently postponed until May 14–15 due to the conflict in Iran. In other words, this Beijing meeting was not held in a stable international environment, but rather was rescheduled after compounding factors such as Middle East war tensions, global oil shipping risks, US domestic inflation, and election pressures.

This point is important. It explains why the issue of Iran was listed alongside China-US economic and trade issues, and the Taiwan issue, as public topics in this meeting.

To What Extent Has the Itinerary Been Disclosed?

As of the morning of May 14, 2026, the publicly available itinerary can generally be summarized as follows:

Date	Publicly Announced Schedule	Source of Information
Evening, May 13	Trump arrives in Beijing; Han Zheng picks him up at the airport	Xinhua News Agency
Morning, May 14	Xi Jinping holds a welcome ceremony outside the East Gate of the Great Hall of the People	Xinhua Quick News
Daytime, May 14	Consultations between both sides at the Great Hall of the People	Reuters citing White House itinerary
Daytime, May 14	Joint visit to the Temple of Heaven	Reuters, AP
Evening, May 14	State Banquet	Reuters, AP
May 15	Tea gathering and working lunch, followed by Trump’s departure from Beijing	Reported by Reuters and AP, citing the White House

There are two details here that are worth noting.

Firstly, the Temple of Heaven was not a randomly selected attraction. Foreign media mentioned that the Temple of Heaven corresponds to ancient Chinese imperial ceremonies for praying for good harvest, and it carries very strong symbolic meaning. For Beijing, this arrangement is both diplomatic and narrative design.

Secondly, scheduling an informal tea gathering and a working lunch on May 15th suggests that this meeting was not simply “a quick meet-and-greet for photos before going their separate ways,” but rather provided both sides with dedicated time to refine the details.

Domestic View: First Steady, Then Discuss Differences, But the Red Line Will Not Loosen

The general consensus in China isn’t really complicated.

The two core phrases from the official narrative are “stability” and “certainty.” Both the Ministry of Foreign Affairs press conference on May 11 and Xinhua News Agency’s commentary on May 12 emphasized the strategic guiding role of head-of-state diplomacy in US-China relations, stressing that amidst global instability, China and the US need to provide more stable expectations for the world.

But this is not merely unilateral goodwill. Domestic statements also carry another very hard line: cooperation can be discussed, but principles cannot be traded away. The Taiwan issue, in particular, is explicitly placed in the position of the “most important issue” regarding China-US relations. Xinhua News Agency’s commentary even re-establishes the One China principle and the three Sino-US joint communiqués as foundational political bases. The meaning is clear: discussing cooperation is acceptable, but do not expect Beijing to be ambiguous regarding its core interests.

If we push domestic public opinion a bit deeper into the social sphere, the emotions are also quite realistic. Beijing respondents interviewed by Reuters were skeptical about Trump’s genuine intentions while simultaneously hoping for “some good policies to emerge.” This type of reaction is very typical; it is not passionate optimism, but rather a desire for things to stabilize and avoid further disruption.

My understanding is that most domestic perspectives viewed this meeting not as the starting point for a comprehensive improvement of China-US relations, but rather as a necessary damage control or de-risking action. Simply maintaining stability is already considered effective; securing some practical progress at the economic and trade level would be an added bonus.

International View: The spectacle will be large, the breakthroughs minor, but both sides cannot afford to miss this meeting

International viewpoints are somewhat more divided, but the main thread is also very clear.

Viewpoint A: Strong emphasis on ritual and form, but limited substantial breakthroughs

AP’s judgment is quite direct: this meeting will likely be “highly visible,” but there probably won’t be any decisive breakthroughs on hard issues such as trade, Taiwan, or Iran.

This isn’t difficult to understand. Because what can be discussed right now mostly involves managing risks, extending ceasefires, and exchanging limited interests; for truly difficult structural conflicts, such as high-tech restrictions, Taiwan arms sales, and supply chain security, no party has backed down enough to easily sign a major deal.

Another View: Trump Needs This Meeting More Than He Did in 2017

Riot’s analysis is sharper. Its judgment is that the balance of power this time is different from 2017. Back then, China appeased Trump with high-spec reception and purchases; this time, it seems more like the US actively admitting that China, as an opponent and trading partner, cannot be avoided.

The reasons behind this change are also not hard to see:

Trump needs economic and trade achievements, especially areas like agriculture, Boeing, and energy—things that can quickly translate into political victories.
The Iran conflict has hurt his approval ratings while also fueling U.S. domestic inflation and increasing pressure ahead of the midterms.
Many members of the American business community are part of this delegation, but their demands are very specific. They aren’t there to hear grand rhetoric; they are there to secure market access, regulatory approvals, supply chain recovery, and easing regulations.

In other words, the White House’s visit to Beijing this time is not just for strategic posturing, but also for seeking practical gains.

Another layer of colder calculation: neither side wants to lose control

This is the one I agree with more.

The issue between China and the U.S. is not that they lack contradictions, but rather that they have too many. This makes it essential to maintain top-level communication. Because once talks between heads of state break down, several issues—such as economy/trade, Taiwan, technology, and regional security—could cascade into one another, eventually becoming unmanageable for anyone.

Therefore, what is truly important about this meeting is not necessarily how many deals are signed, but whether both sides have the capacity to manage conflict zones and set aside/isolate issues that cannot be resolved immediately.

What exactly prompted this meeting?

If I had to summarize it in one sentence, I would say: It was common pressure, not mutual affection, that led to this meeting.

More specifically, it is four forces acting simultaneously.

1. The ceasefire following Busan requires extension

After the Busan meeting in October last year, the two sides at least temporarily prevented their economic and trade ties from spiraling out of control. This temporary balance was inherently fragile, and without continued talks, it could deteriorate again. The Beijing meeting first and foremost maintained this truce/stability.

2. Both Parties Have Specific Transactions to Discuss

What the US wants to talk about is very practical: agricultural products, Boeing, energy exports, market access for US companies in China, and regulatory issues encountered by chip and AI companies.

What the Chinese side wishes to discuss equally involves whether restrictions on chip equipment and advanced semiconductors can be eased, if the pace of technology containment can slow down, and whether economic and trade cooperation can return to a more predictable track.

This is not a conversation about values; it’s an exchange of terms (or: it’s conditional bargaining).

3. The Iran War Puts the US and China on the Same Table Again

The Iranian issue was not a peripheral topic to Sino-US relations; it almost became an external forcing variable during these talks. The US hopes that China can utilize its influence in energy and regional affairs to help de-escalate the situation. Meanwhile, China is unwilling for global energy and shipping risks to continue spiraling out of control, further impacting foreign trade and the global economy.

Both sides hold differing positions, but neither wants to see the situation continue to deteriorate. This is enough motivation for a meeting.

4. Both sides both require a relatively stable `2026`

China is preparing to launch the “15th Five-Year Plan” and will also be hosting APEC. The US has an independent 250th anniversary narrative, alongside pressures from the G20 and mid-term elections. Neither side lacks strong statements, but both desperately need a predictable external environment.

Therefore, the true significance of this meeting does not lie in who convinces whom, but rather in both sides realizing that further delay would entail higher costs.

My Judgement

This Beijing summit is unlikely to usher in a “new era” for Sino-US relations. What it is more likely to deliver, however, is a limited but significant outcome: having both sides reaffirm that they will continue competing, but within defined boundaries.

This doesn’t sound romantic, but it is very realistic in the current Sino-US relationship.

If there are any actual results later, the most likely things to be implemented first will not be grand narratives, but rather several types of pragmatic actions: continuing to maintain a ceasefire in economic and trade relations; releasing small signals regarding purchases or approvals, giving the business community some progress they can report externally; while also leaving some room for maneuver concerning the issues surrounding Iran and Taiwan before completely closing off all options.

Regarding deeper competition—especially concerning high technology, global value chains, and security issues—these matters will not simply vanish after this meeting. They are merely being placed into a more manageable rhythm for the time being.

How should I put it, this meeting is not about proving mutual trust between China and the U.S., but rather about demonstrating whether both sides still have the capacity to continue engaging despite deep mistrust.

This is its true weight.

References

Ministry of Foreign Affairs of the People’s Republic of China: President Xi Jinping Speaks with U.S. President Donald J. Trump on the Phone (2026-02-04)
Ministry of Foreign Affairs of the People’s Republic of China: Foreign Ministry Spokesperson Guo Jiakun’s Regular Press Conference on May 11, 2026
Ministry of Commerce of the People’s Republic of China: MOFCOM Spokesperson Responds to Reporters on Issues Related to China-US Economic and Trade Consultations (2026-05-10)
[Xinhua Net: China and the U.S. Hold Economic and Trade Consultations in South Korea (2

Writing Notes

Original Prompt

The recent summit between China and the U.S.: Compile the background details and itinerary of the related events. Discuss how both domestic and international parties view this meeting, and analyze the main factors that led to it.

Summary of Writing Approach

Brainstorming Expansion

Direction	Inclusion in Main Text	Reason
Comparing differences in protocol specifications from Trump

The big model development has indeed drawn the internet giants into the same competitive arena.

Tue, 12 May 2026 00:59:12 +0800

My previous article covered the semiconductor cycle, and I feel like there’s a piece of background/context missing.

Your judgment/conclusion regarding this point—the overall direction is correct. Furthermore, I believe it is a prerequisite that is easiest to overlook when trying to understand this current semiconductor boom.

A more accurate way to put it is not that “all internet giants are fighting in the same field,” but rather: Large Models have, for the first time, brought together major players previously scattered across different domains—such as search, advertising, social media, e-commerce, office productivity, cloud computing, and content distribution—into direct competition within the same technical stack.

This technology stack includes models, computational power, inference, cloud, Agents, distribution gateways, and commercialization closed loops. Everyone’s original “moat” is different, but now we must all fill the same gap. Those who fail to do so will see their future search entry points, ad pricing, office suites, e-commerce conversion, and social traffic distribution rewritten by others.

Why is your viewpoint basically valid

Previous internet competition was more like everyone staying within their own established territories (or: was more siloed).

Google / Baidu primarily focus on search.
Meta / Tencent primarily focus on social networking and traffic distribution.
Amazon / Alibaba primarily focus on e-commerce and merchant ecosystems.
Microsoft focuses on office software and enterprise software.
Although AWS, Azure, and Google Cloud are involved in the fight, that is more of a battle over cloud infrastructure and enterprise IT.

It’s different now.

Large models are not a single function; they are more like an “master switch” that will reverse-engineer/supersede all entry points. Search can be rewritten, ad placements can be rewritten, customer service can be rewritten, code generation can be rewritten, e-commerce recommendation can be rewritten, and enterprise knowledge bases can also be rewritten.

Therefore, although these giants are still making money in their respective established domains on the surface, fundamentally they are all competing for the same thing:

\[ \text{AI Competitiveness}=\text{Model Capability}+\text{Computational Resource Supply}+\text{Distribution Channels}+\text{Commercialization Closed Loop} \]

The difference merely lies in who is superior in which aspect.

Microsoft is strong in enterprise distribution and Azure.

Alphabet is strong in search, advertising, and proprietary model stacks.

Amazon is strong in AWS, chips, and enterprise cloud customers.

Meta is strong in traffic entry points and advertising scenarios.

Tencent is strong in super apps, gaming, and advertising implementation.

Ali is strong in e-commerce, cloud, and industrial clients.

Baidu is strong in search, AI Cloud, and ERNIE.

So your judgment is correct, but we need to add one caveat: Everyone is fighting on the same large battlefield, not using the same set of weapons or competing for the same group of users.

This is also why semiconductors are linked/grouped together

The previous text discussed storage and South Korean semiconductors; at that time, I focused on specific product categories like HBM, DDR5, and eSSD.

However, looking at a higher level, what truly fueled all these developments was the way the capital expenditure of internet giants began to concentrate in one direction.

It’s not that Google needs to buy a little more server capacity this year, but Meta might follow suit next year.

Instead, during the 2025 to 2026 phase, Microsoft, Google, Amazon, Meta, Alibaba, Tencent, and Baidu are almost all focusing on AI infrastructure, model training, inference services, Agent platforms, and AI distribution entry points. While their specific terminology varies, money is pouring into areas like GPUs, HBM, networking, SSDs, data centers, and power.

This is uncommon in internet history.

In the mobile internet era, many companies are involved, but not every single one needs to build its own operating system.

The short video era is booming, but not every company needs to build and train its own foundational recommendation model infrastructure.

The era of cloud computing has been long, but traffic platforms and gaming companies like Meta and Tencent have not all shifted their focus to the cloud.

Large models are something that almost every platform giant believes cannot be missed this time. This is what makes this semiconductor cycle different from previous ones.

What Is the Core Area of Competition Among All Players Right Now?

Let’s first clarify the definitions/scope. (Alternative options depending on context:)

We need to establish the parameters first. (If discussing data or technical metrics)
Let’s get aligned on the basis first. (If referring to an agreement or overall strategy)

For the table below, revenue and net profit should use the official disclosures based on each company’s latest completed fiscal year as of May 12, 2026. However, regarding the “AI Investment” column, disclosure methods vary greatly among companies; some report Capex, others provide a three-year investment plan, some report R&D figures, and some only provide quarterly expenditures.

So this column can only be used to view the strength and weakness direction, not for mechanical ranking.

Company	Existing Cash Cow / Core Business	Current AI Leverage Points	Latest Full Fiscal Year Revenue	Latest Full Fiscal Year Net Income	Most Recently Disclosed AI Investment Scope
Microsoft	Office, Windows, Enterprise Software, Azure	Azure + OpenAI Ecosystem + Copilot + Enterprise Agent	FY2025 Revenue $281.7 Billion	FY2025 Net Income $101.8 Billion	Approx. $80.1 Billion in CapEx for the first 9 months of FY2026; AI business annualized revenue reached $37 Billion in Q3 2026
Alphabet	Search, Ads, YouTube, Android, Cloud	Gemini + Search AI + TPU + Google Cloud	FY2025 Revenue $403.0 Billion	Total GAAP Net Income for four quarters of 2025 was approx. $132.2 Billion	FY2025 CapEx $91.4 Billion; FY2026 guidance $175B–$185B
Amazon	E-commerce, AWS, Advertising	Bedrock + Trainium/Inferentia + AWS AI infra + Nova	FY2025 Revenue $716.9 Billion	FY2025 Net Income $77.7 Billion	Approx. $147.3 Billion in CapEx over the last 12 months as of March 2026; Over 2.1 million AI chips delivered in the past 12 months
Meta	Facebook / Instagram / WhatsApp Advertising	Llama + Ad Recommendation + Meta AI + Smart Glasses / Agent	FY2025 Revenue $200.97 Billion	FY2025 Net Income $60.46 Billion	FY2025 CapEx $72.2 Billion; FY2026 guidance $125B–$145B
Tencent	Gaming, Social, Advertising, Fintech	Hunyuan + WeChat /

The most notable thing about this table is not who makes the highest profit, but rather who possesses enough robust “cash cow” resources to withstand the upfront investment period required for AI.

Microsoft, Alphabet, Amazon, and Meta—these four companies are all essentially printing money while expanding production.

Tencent and Alibaba are similar; their old businesses are still providing blood, so they can ramp up their AI investments.

Baidu’s problem is more apparent: they bet early, but their scale and cash flow depth are weaker than the previous few companies, making them naturally less resilient under pressure.

Although all are involved in the AI craze, each company’s approach is actually different

Upon closer examination, it can be seen that although these companies are all engaged in intense competition, their strategic positioning is different.

Category 1: Selling the Shovels While Also Entering the Field

Microsoft, Alphabet, and Amazon belong to this category.

They have cloud, chips or accelerators, enterprise customers, and models or model ecosystems.

What is most frightening about these companies is that, for them, AI isn’t a standalone product, but rather an upgrade tax on the entire platform.

You want to train a model, so you need to buy cloud services.

You need to infer, you need to rent a GPU.

You will need to become/develop an Agent, and you will need to purchase platform services.

You need to provide employees with AI tools, which requires purchasing related capabilities for Copilot, Gemini, and Bedrock.

So, they don’t just want to create a blockbuster AI App; they aim to integrate AI into every part of the entire IT budget (or: IT expenditure).

Type Two: The Traffic King – Optimize Your Distribution Channels First

Meta, Tencent, and Baidu are closer to this category.

What they value most are not enterprise contracts, but rather traffic funnels/entry points, advertising systems, content ecosystems, and high-frequency user time spent.

Meta is the most typical example. It doesn’t need to sell AI to all enterprise clients; merely by applying AI in ad targeting, content recommendation, creative generation, and conversational entry points, it is enough to enhance advertising efficiency and commercialization capabilities.

Tencent is similarly positioned. Areas like WeChat, advertising, gaming, and cloud are inherently natural use cases for embedding AI. While it doesn’t necessarily have to compete with Microsoft by tackling the entire Office suite, it will certainly focus on dominating AI assistants within WeChat, improving ad placement efficiency, enhancing game content production, and deepening collaboration in enterprise WeChat.

Baidu, on the other hand, is more like a traditional search company striving to make a turnaround. It has models, cloud capabilities, and search, but it also faces the greatest pressure from its old advertising business. Therefore, it needs to both maintain its traditional traffic base while simultaneously boosting AI commercialization.

Category Three: Ecological King, Aiming to Transform AI into Trading and Industrial Infrastructure

Alibaba is the most typical example among this category.

Its ambition is not merely to create Qwen, nor is it simply about divesting Alibaba Cloud. What it truly seeks is to integrate AI throughout the entire ecosystem of e-commerce, merchant tools, customer service, marketing, supply chain, and industrial cloud.

So Alibaba has repeatedly emphasized “AI + Cloud” in recent years. These four words are not just a slogan because it realized earlier than many companies that: the model itself might not be the most profitable, but the model will redefine cloud and transaction platforms.

Why Has This Conflict Concentrated Semiconductor Demand So Heavily

Seeing this, we can actually get back to semiconductors.

If only a single company is optimistic about AI, what it generates will primarily be themed investing (or thematic investments).

If there are seven or eight platform companies with the strongest cash flows, and simultaneously converge capex, R&D, models, inference, Agents, and distribution channels onto the same layer/level, then it will no longer be a theme, but actual orders (or: real demand).

But in reality, the final orders will turn into these things:

GPU and customized AI chips.
HBM and high-end DRAM.
Enterprise-grade SSD and higher bandwidth storage.
High-speed networking, switching chips, optical modules.
Data centers, power, cooling.

This is why when analyzing the semiconductor sector lately, you shouldn’t only focus on a single memory manufacturer or a single South Korean company.

The true background is: Internet giants have unusually ramped up their efforts simultaneously in the same arms race.

Conclusion

As of May 12, 2026, my assessment of your point is:

Basically correct.
A more accurate statement should be: internet giants are not completely engaged in “the same product track,” but rather in an arms race over the same foundational AI capabilities.
The importance of this matter is not just because AI will produce blockbuster applications, but because it is drawing originally scattered profit pools—such as search, advertising, social media, e-commerce, office work, and cloud services—back into one common underlying competitive framework.
This is also one of the core backgrounds explaining why the semiconductor industry, especially the chains involving GPU, HBM, storage, and data centers, has been so intense recently.

Well, how do I say it? Previously, everyone could maintain their own separate domains. That’s not feasible anymore. The matter of large models has evolved from merely a new trend into an infrastructure war that platform companies cannot afford to exit.

References

Notes on Writing

Original Prompt

The previous text mentioned high AI demand and surging semiconductor prices, but there is a background that is easy for everyone to overlook: large AI models. This is rare—it’s a field where all major internet giants have started heavily competing. Previously, they developed in their own respective domains; it is uncommon now for them all to be intensely vying within a single sector. First, can I confirm if my viewpoint is correct? Then, let’s outline the sectors each company is developing, along with their respective revenues, net profits, and investments in AI.

Writing Idea Summary

First, judge whether the point of view is valid, and then clarify “to what degree it is valid,” avoiding making sweeping statements with one sentence.
In the main body, rewrite “same industry track” as “same layer of technical stack,” as this better aligns with the true positioning of various companies.
In the middle, use a summary table to place cash cows, AI entry points, revenue, net profit, and AI investment metrics side-by-side, allowing readers to quickly survey the whole picture.
The article intentionally writes out that “AI investment cannot be directly compared,” to avoid forcing rankings using disparate disclosure metrics.
This piece deliberately does not include OpenAI, Anthropic, xAI, and ByteDance together in the table, as they are non-public or have varying disclosure methods, which could easily bias the comparison.

Expanded Brainstorming

Area	Whether to include in the main text	Reason
Including Apple in the comparison?	No	It is certainly affected by AI, but it is not a typical sample within this

The thing that collapsed around the May 1st period wasn't the box office, but the trust.

Tue, 05 May 2026 08:46:19 +0800

The movie box office during this year’s May Day holiday was disappointing; it’s no longer simply a matter of “which film failed to be a massive hit.”

If we only discuss the economic downturn, it can certainly explain some things. People are not as financially comfortable as they used to be, so spending money is more cautious—that is a reality. However, I feel that blaming all the problems on the economy is absolving the industry of its fault. It’s not that audiences suddenly stopped liking movies; rather, their patience has been depleted by round after round combination of high promotion/marketing efforts, strong screening slots, and weak content.

Look at the data first.

Year	Period Scope	Total Box Office	Viewership	Notes
2021	May 1st to May 5th	1.668 billion RMB	44.1054 million	According to the China Film Administration, set a record for the May Day period at that time.
2022	April 30th to May 4th

The year 2022 cannot be directly used as proof that the movie market has collapsed because that year had a pandemic and cinema operational restrictions, making it an outlier. What is genuinely concerning is this line from 2024 to 2026.

In 2023 and 2024, for two consecutive years, it remained around 1.5 billion, with the number of viewers also exceeding 37 million. By 2025, the box office directly dropped to 747 million, and the number of viewers only reached 18.895 million. As of the evening of May 4, 2026, it had already exceeded 660 million, and it will continue to increase on the last day, but even with a final boost, it is difficult to return to the level of 2023 and 2024.

This is not normal volatility; it is a clear discontinuity/break.

What’s more embarrassing is that the average ticket price for the May Day period in 2026 dropped to around ¥36.9, which is the lowest during this time over the past four years. The tickets are no longer that expensive, but cinema attendance has not significantly recovered. This suggests that the problem is not merely “movie tickets are too expensive,” but rather that many people have completely lost the impulse/urge to “go see a movie at the cinema.”

It’s not like there were never any bad movies before. / There were actually bad movies before.

In past years, audiences haven’t been unable to pay for mediocre films. It was just that at the time, movie-going as a holiday activity carried a certain inertia. During Spring Festival, National Day, and May Day, people saw it in their social circles, short video platforms promoted it, theaters were fully booked, and marketing buzz was everywhere. Many people bought tickets not because they were truly moved by the film, but because “it seemed like everyone else was watching it.”

This kind of business is very effective in the short term.

Trending stars can drive the first wave of pre-sales, emotional marketing can turn controversy into buzz, and platform algorithms can push clips to every single person. As long as it takes off during the opening weekend, the box office numbers will become new promotional materials. Even if the film itself is mediocre, it can still maintain the appearance of being a massive hit.

But this approach has a side effect: it depletes the audience’s trust in movie theaters.

I still have no interest in catching up on films like Full River Red or Spicy Rolling Hot. It’s not that they failed commercially; of course, they succeeded. The problem is that when the public discussion surrounding a film increasingly resembles a marketing campaign, it becomes difficult for audiences to believe that what they are seeing is the work itself. Online, people often say that the promotion and marketing costs of certain films exceed the production costs. I cannot find reliable public financial data for this claim, so I cannot state it as a fact. But from an audience perspective, it shows one thing: people have begun to suspect that the money was not spent on making the movie, but rather on making you feel like you must watch it.

Once such suspicion has formed, it is difficult to repair it by relying on the next hot search.

In the past, even after criticizing a bad movie, people might still go see it during the next screening cycle. That is not the case anymore. Short videos, streaming series, games, concerts, travel, camping, city walks—there are too many choices for holidays now. If movies cannot provide a reason that they must be watched in a cinema, they will naturally get crowded out/marginalized.

Economic downturn is only an accelerator, not the sole reason.

During times when money is tight, people won’t stop consuming completely; rather, they will re-prioritize their spending. In the past, spending a couple hundred [currency unit] on a movie ticket plus drinks and popcorn was acceptable. Now, that same amount of money can cover a meal, buy a game, or afford an outing. If movies still rely on hype, traffic, or emotional manipulation to compete for this spending power, the audience will naturally be more hesitant.

What is more problematic is that in the past few years, the film market has led audiences to develop a defensive consumption psychology: the harsher the pre-release hype, the more they tend to wait; the denser the trending topics, the more artificial it seems; and the more exaggerated the opening day box office, the less it indicates if the movie is truly good. Audiences are not ignorant; they simply were too lazy to scrutinize things before.

This phrase is quite contextual, and the best translation depends on whether the speaker means “keeping account/score” in an argument, or simply becoming overly critical of details.

Here are a few options based on different implied contexts:

1. (Most common - Implies arguing or scoring faults):

You’re starting to keep score now. / Now you’re getting picky.

2. (Implies criticizing small details - Nitpicking):

You’ve started nitpicking. / Now you’re fussing over the details.

3. (More general/literal, but less natural in English):

Now it’s starting to be scrutinized.

The collapse of blockbuster seasons like the May Day period isn’t that audiences suddenly betrayed movie theaters; rather, it signifies that the guarantee that cinemas once offered has evaporated. Previously, buying a ticket felt like participating in a holiday ritual; now, it feels more like making a calculated risk. If the film is bad, you lose your money, you waste your time, and you are still forced to endure two hours of awkwardness.

So I don’t think it’s a bad thing.

In the short term, both cinemas and production companies are struggling. In the long term, if audiences rely less on blind following and the market has less artificial hype, things might actually return to normal. Promotion can lure people into the cinema once, but it cannot make them continuously trust the cinema. What can truly bring people back is the content itself.

The problem is whether the industry is willing to acknowledge this.

If they continue to point fingers at economics, ticket prices, weather, and short videos, and keep filling up the screening slots with popular stars and marketing platitudes, then the May Day period is only the beginning. The audience won’t formally announce their departure from movie theaters; they will just stop buying tickets time after time.

This is more frightening than curses/swearing.

References

Writing Notes

Original Prompt

$blog-writer Everyone is saying that this year’s May Day film box office has collapsed. Comparing it to previous years’ data, it is indeed dismal. Putting aside the economic downturn and the fact that people aren’t as wealthy in their pockets anymore, are previously released flop films not at fault? The spending driven by capital + hype will eventually backfire. This is the best backlash—are people simply not going to movie theaters? Man Jiang Hong and Hot & Spicy, I still can’t bear watching these two movies; their promotional costs might even exceed their production costs. Remember to organize a specific table comparing how the May Day box office has collapsed over these five years.

Summary of Writing Ideas

Create a table using May Day period data from 2021 to 2026, but clearly state that 2026 is not yet the final closing number.
Mark 2022 separately as an anomaly due to the epidemic, and focus the main text on the gap/discontinuity between 2024 and 2026.
The core argument is that the audience’s trust in high marketing expenditures, guaranteed screening slots, and low-quality content has begun to face a backlash.
Do not treat “marketing expenditure exceeding production costs”

Huawei is blocked, but Xiaomi can find TSMC

Mon, 04 May 2026 15:04:48 +0800

I recently revisited this issue, and the conclusion is quite straightforward: It’s not that Huawei “cannot use TSMC”; rather, it is that U.S. regulations have severely constrained its compliant pathways. Xiaomi is able to access TSMC simply because it was never included in the same export control list as Huawei.

Why Has Huawei Been Restricted?

Huawei was first placed on the Entity List by the U.S. Department of Commerce on May 15, 2019. This list means that exporting American technology to this company requires a license, and that license can also be rejected.

The real roadblock was the rules upgrade on May 15, 2020. The U.S. Department of Commerce explicitly stated that it would restrict Huawei’s ability to use American technology and software overseas for chip design and manufacturing, effectively sealing off the path of “finding an overseas foundry to bypass it.” TSMC made this even clearer in its 2020 annual report: to ensure compliance, they stopped shipping to Huawei starting from September 15, 2020.

Therefore, Huawei is not merely restricted from “buying chips,” but is bottlenecked across the entire value chain encompassing design, manufacturing, contract fabrication, equipment, and software. You can understand it this way: for Huawei, the problem isn’t simply finding a factory, but whether that factory dares to accept the work, and if doing so would violate American regulations.

Why Xiaomi Still Can Work with TSMC

Xiaomi and Huawei do not have the same treatment. Xiaomi has not been placed on the Commerce Department’s Entity List, so it is not in the situation where “any contract taken by the factory could potentially violate restrictions,” like Huawei.

Many people mistakenly equate Xiaomi and Huawei regarding events in 2021, but this is inaccurate. Back then, Xiaomi was listed once by the U.S. Department of Defense as a “Chinese military company,” but that restriction applied to investments made by U.S. persons. The court issued a preliminary injunction in March 2021, and the U.S. officially lifted this securities restriction on May 25, 2021. It was not an export embargo like Huawei faced, nor was it a comprehensive ban on chip foundry services.

More importantly, in recent years, the US has become increasingly targeted. Its restrictions are not aimed at all Chinese consumer electronics, but specifically at high-end AI chips, supercomputing, and related equipment. When updating the rules on October 17, 2023, BIS stated it very clearly: the focus is on advanced computing semiconductors, semiconductor manufacturing equipment, and supercomputing items, with the core objective being military AI capabilities and the risks associated with military-civil fusion.

This explains why Xiaomi’s XRING O1 can still utilize TSMC. It is a consumer electronics chip, not a highly restricted AI training chip like Huawei’s. As long as the customer has not been blacklisted and the chip specifications do not cross regulatory red lines, TSMC still maintains compliance scope.

Who Else Seeks TSMC Foundry Services

Yes, and quite a few. Public reports already mentioned in 2024 included Chinese AI chip companies such as MetaX and Enflame, which had to downgrade their designs before resubmitting them merely to secure TSMC capacity. This detail speaks volumes: the US is not restricting whether “Chinese products can use overseas foundries,” but rather ensuring that “high-end AI chips cannot easily bypass (these restrictions).”

In other words, what is truly being heavily restricted is the AI chip line, not all Chinese companies or all Chinese products. Consumer-grade SoCs might still have some room, but high-end AI accelerators and training chips are becoming increasingly difficult to obtain.

How Was Huawei’s AI Chip Discovered?

The link/chain this time has been exposed not because of a public admission from Huawei, but through a device teardown.

In October 2024, Reuters reported that TechInsights disassembled a Huawei product and found a chip that appeared to be manufactured by TSMC; subsequently, TSMC also notified the US side. Immediately after, Reuters further reported that TSMC stopped shipping to Sophgo, a Chinese chip design company, because it found its fabricated chips in Huawei’s AI processor.

The product corresponding to this line is Huawei Ascend 910B. In other words, what the outside world is building/assembling is not a phone chip, but rather Huawei’s AI processor. The reason it is problematic is not merely because of “which foundry was used,” but because it falls into one of the most sensitive categories of chips in the United States.

My Own Assessment

Upon prolonged observation, it becomes clear that America’s strategy is not about generalized efforts but rather a progressively targeted approach. Huawei serves as the benchmark; AI chips are the main battleground, while foundries are merely an execution component.

Xiaomi being able to use TSMC does not mean that US restrictions on China’s chips have eased; it just means they haven’t crossed the next, narrower, and harder threshold. Huawei’s difficulties are not because “Huawei is bigger,” but because it has been placed squarely in the crosshairs of export controls—it cannot circumvent them.

References

Writing Notes

Original Prompt

Please query relevant government documents and explain why TSMC cannot manufacture Huawei’s chips. Regarding Xiaomi’s “Xuanjie” chip, if TSMC can manufacture it, wasn’t Xiaomi sanctioned by the US? Besides Xiaomi, are there any other domestic companies using TSMC for manufacturing? The focus now seems to be heavily on locking down AI chips. Could you detail the specifics of Huawei’s AI chip outsourcing/manufacturing: how was this discovery made, and what specific product is involved?

Summary of Writing Ideas

This piece focuses on the rule differences between “why Huawei is blocked vs. why Xiaomi can still operate.”
The section on Huawei integrates historical timelines: listing names in 2019, and adding the foundry chain in 2020.
The section on Xiaomi clarifies separately: it encountered investment restrictions, but not a Huawei-style export ban.
The article deliberately avoids developing into a comprehensive overview of the US-China tech war, only retaining parts directly related to TSMC’s foundry services.
Huawei’s AI chip discovery process is structured as a single line following “teardown - tracing - foundry chain.”
The conclusion returns to a practical judgment: what is currently targeted are high-end AI chips, not all Chinese chips.

New Drone Regulations Take Effect, DJI First Gets Blocked in Beijing

Mon, 04 May 2026 14:58:12 +0800

After May 1st, when regular people buy a drone, it feels like they are getting more than just a flying camera.

It is more like a flying terminal equipped with identity tracking, trajectory monitoring, and approval mechanisms. The impact of this change on DJI is not simply selling fewer machines; rather, the entire product definition of consumer drones has been rewritten. What DJI can actually do is limited. At least in heavily regulated areas like Beijing, it can only reclaim its distribution channels and then integrate compliance capabilities into both its products and service workflows.

This Change Is Not a File

What is easily confused are policies at two different levels.

At the national level, two mandatory national standards will be implemented starting on May 1, 2026. One is the “Real-name Registration and Activation Requirements for Civil Unmanned Aerial Vehicles,” addressing the question of “who can fly.” UAVs must complete real-name registration and pass system verification before they can activate and obtain flight capability. The other standard is the “Standard Specification for System Operational Identification of Civil Unmanned Aerial Vehicles,” addressing the question of “who is flying.” According to interpretations by the Civil Aviation Administration, civil UAVs that lack operational identification sending functions should cease operation starting from May 1, 2026.

It’s no longer as simple as just affixing a real-name registration QR code like before. Previously, the general user intuition was that simply registering it after purchase would be enough to prevent misuse. Now, the regulatory logic is much more advanced: whether the machine can be activated, whether it can be identified while in flight, and whether the manufacturer provides interfaces and upgrade plans—all of these factors will become integral parts of the product itself.

At the Beijing level, it is more direct.

The “Beijing Regulations on the Management of Unmanned Aerial Vehicles” will also be implemented on May 1, 2026. Beijing has designated the entire administrative area as a controlled airspace for unmanned aerial vehicles, requiring an application for all outdoor flight activities. More critically, sales, transportation, carrying, and storage are all subject to control: It is prohibited to sell or rent unmanned aerial vehicles and their core components to units or individuals within the administrative region of Beijing, and transporting or carrying drones and their core components into Beijing is also forbidden. The situation concerning existing drones that complete real-name registration and information verification will be addressed separately.

Therefore, what the outside world perceives regarding “DJI’s withdrawal” (or discontinuation) is not that sales are suddenly prohibited across the entire national market, but rather that the specialized Beijing market has been individually restricted. This distinction is crucial. National policy focuses on traceability requirements, whereas Beijing’s policy enforces high-intensity spatial control.

What is the Impact on DJI?

The most direct impact of DJI is that the Beijing consumer market has essentially lost normal retail scenes.

Jiemian News reported that after the afternoon of April 29th, DJI stores in Beijing will no longer sell drone products, and online platforms have also ceased shipping to the Beijing area. If Beijing users require subsequent repairs, the stores will not bear repair responsibilities; they must instead use self-mailing repairs from other provinces/cities or seek replacements through authorized service centers. This change is quite problematic for a consumer electronics company because it not only means losing sales in one city but also impacts the entire customer experience chain.

Drones are not cell phones. While a phone may be difficult to buy in Beijing, you can purchase it elsewhere, and it will function the same way when brought back. Drones are different; buying, carrying, flying, storing, or maintaining them might trigger the same set of regulatory controls. Even if Beijing users have the desire to purchase [drones], they are often discouraged by practical concerns such as, “Can I bring it back after purchasing?”, “Is flying allowed?”, and “What happens if it breaks?”

The second impact is the continued rise in product compliance costs.

DJI’s previous strength was transforming aerial drones into very mature consumer electronics products. The process—unboxing, activating, connecting the App, taking off, and shooting—became smoother and smoother. However, after new regulations, product experiences must embed more regulatory actions: refreshing real-name registration status, interacting with the UOM platform, applying for flight activities, reporting operational identification data, and explaining the relationship between no-fly zones and government controlled airspace.

Users will not perceive these things as “functional enhancements.” Users will only feel that it makes things more complicated/troublesome. However, manufacturers must implement them, and if done poorly, they will become post-sale issues.

The third impact is that DJI’s consumer advantages will be compressed, making industry capabilities more important.

In a place like Beijing, demand for personal aerial photography is severely restricted. The sectors that can genuinely navigate the necessary procedures are typically those related to emergency response, surveying and mapping, inspection, academic research, public safety, or film production—all areas with clear stakeholders and defined purposes. These applications do not necessarily prioritize “cheap and fun,” but rather focus on mission reliability, complete approval documentation, data traceability, and implementable services.

This isn’t necessarily all bad news for DJI. DJI already has product lines dedicated to industrial applications, covering scenarios such as agriculture, energy, and mapping/surveying. However, the potential scope for consumer-grade products will narrow. The narrative of average people buying a Mini and taking it on casual trips is becoming increasingly impractical in highly regulated cities.

How Did DJI Respond?

In the short term, DJI’s response is relatively pragmatic: Beijing channels are removing products from shelves; online sales restrict shipping; after-sales service shifts to cross-regional mailing for repair or authorized service centers; and public documentation continues to guide users toward real-name registration and flight reporting.

This isn’t the most elegant approach, but it is a realistic one. Local regulations have explicitly prohibited the sale and rental of drones and core components to Beijing, leaving the company with little room to maneuver. Continuing to push sales is meaningless; instead, it will transfer compliance risks to the stores and users.

Even on DJI’s official support pages, they have actually been guiding users toward the UOM system for some time now. For example, the real-name registration page prompts DJI drone users to complete their registration at the earliest and mentions that they can check their verified identity status within DJI Fly. The flight reporting page also details procedures such as the UOM application process, pre-flight reporting, and post-flight submission. In other words, it wasn’t that DJI only started facing regulation on May 1st; rather, on this date, many things that were originally merely “advisory” became hard constraints for whether or not the product could continue to be used.

I think this is what makes DJI feel the most discomfort.

It cannot solely focus on hardware, nor can it be limited to just the image experience. It must also function as a compliant middleware layer: when users do not understand policies, the App must explain them; if a user cannot take off/launch, the system must provide reasons; when government platform interfaces change, the firmware and cloud services must adapt accordingly; and when existing devices no longer meet requirements, it must offer upgrade or replacement solutions.

Consumer electronics companies fear this kind of thing the most. Because these types of costs are difficult to reflect in specification sheets, and they are also hard to turn into a selling point that users are willing to pay for.

Future Development of Drones

Drones are highly likely to split into two main paths/tracks in the future.

One type is more like a tool. It serves specific tasks: requiring applications before flight, conducting identification during flight, and generating records after flight. It can be applied to scenarios such as inspection, surveying/mapping, agriculture, fire fighting, logistics, and urban governance. The focus here is not on spectacular shots, but rather on reliable processes, clear accountability, and data closed-loops. If the low-altitude economy truly wants to develop, it should rely on this path, rather than people just flying randomly in city parks.

Another trend is that it’s becoming a more restricted imaging toy. Consumer aerial photography will still exist, but it will become increasingly dependent on permissible flying zones, scenic area rules, local policies, and platform compliance. In the future, people buying drones may need to prioritize the use case—much like when buying a car—rather than focusing solely on battery life, image quality, and weight.

I don’t entirely agree with simply interpreting this situation as “suppressing DJI.” Beijing’s regulations are indeed strict and unfriendly to ordinary consumers. However, from a regulatory perspective, the biggest difference between drones and cameras is that drones enter public airspace. As long as it can fly, carry a camera, and traverse walls and roads, they cannot be regulated simply like standard consumer electronics.

The real problem is that regulation cannot rely solely on prohibition.

If the future only imposes restrictions like “cannot fly, cannot sell, or cannot carry,” then the drone industry will be relegated to specialized equipment used by a select few entities, effectively pushing out ordinary consumers and small business teams. A better direction for development would be to clarify the compliance procedures: specifically, which areas are permissible for flight, how long is the application process, why might an application be rejected, how to register temporary activities, how cross-city transport should be handled, how existing equipment can be upgraded, and how liability boundaries are determined.

DJI must continue moving in this direction. While making the hardware smaller, more stable, and safer is only one part, what is more important is productizing features such as identity recognition, geo-fencing, operational reporting, flight application processes, and after-sales workflow management. It might not sound glamorous, but it could be more critical than further improving image quality.

Drones have transitioned from free operation to institutional control—this shift has already occurred. For DJI, Beijing was merely the first city where this reality was made public. For ordinary users, in the future, asking “Where can I legally fly?” before buying a drone may be more important than asking “How clear are the shots this machine takes?”

References

China Civil Aviation Administration: “Interpretation | Mandatory National Standard ”》
China Civil Aviation Administration: [“Interpretation | ”](https://www.caac.gov.cn/XXGK/XXGK/ZCFBJD/202601/t20260


Notes on Writing
Original Prompts

$blog-writer The impact of mainland China’s May 1st drone new policy on DJI, how did DJI respond? How should drones develop subsequently?

Summary of Writing Ideas

This piece breaks down the changes on May 1, 2026, into national mandatory standards and Beijing local regulations, preventing the mistake of listing the Beijing sales ban as a nationwide one.
The core conclusion is that drones are shifting from consumer electronics products to regulated flight terminals, meaning DJI’s impact is not just on sales volume, but also requires rewriting product experience and service linkages.
The focus for DJI’s response should be writing about channel delisting, shipping restrictions, after-sales migration, and UOM compliance guidance, rather than framing it as proactive marketing actions.
Future developments are leaning towards “toolification” and “compliance productization,” without expanding on the side branch of overseas market limitations and US security reviews.



Filming a short drama about zombies and spiritual beasts—the first thing that changes (or gets cut) is the budget spreadsheet.
Mon, 04 May 2026 13:57:52 +0800
The most interesting thing about AI short dramas is not that they can immediately replace live-action short dramas, but that they turn genres that were previously considered “too risky to bet on” into viable subjects for experimentation.
Zombies, armies, spirit beasts, fantasy—these elements are incredibly difficult to execute in traditional live-action short dramas. It’s not a failure of screenwriting; it’s that every single step requires funding: extras, costumes and makeup, sets, special effects, staging/choreography, safety measures, and post-production. Furthermore, the commercial logic of short dramas demands speed, low cost, and high-frequency uploads. The greater the imaginative scope of a theme, the easier it is for costs to spiral out of control.
AI first revised this budget sheet. It doesn’t guarantee that every shot will be high-end, but it manages to constrain elements that previously required building sets, hiring massive crews, or doing elaborate special effects, down to a feasible/experimental scope. Zombies no longer need to organize 100 extras; mythical creatures don’t automatically drain a cinematic-level VFX budget; and army scenes don’t necessarily have to start with costumes and locations.
In the era of live-action short dramas, many genres aren’t that no one wants to film them; rather, producers are simply afraid to start filming.
If you want to film an army scene, you must first address where the personnel come from, how to stage the environments, and how to manage the vehicles and explosions. Filming zombies requires dealing with makeup effects, chases, crowd scenes, and special effects. Filming mythical beasts or high fantasy is even more complicated because it is difficult to verify preliminary elements like “what exactly does this thing look like, and will its movement be believable?” on a low budget.
So, in the past, many projects died very early on. They weren’t killed during broadcast, nor were they rejected by audience feedback; they died at the production table. Before we even entered proper creative discussions, the budget would already shut it down and tell you to stop dreaming.
What AI first contributed to short-form dramas was establishing this threshold. It’s not about how much it elevated the aesthetic ceiling of these dramas; rather, it is pulling a batch of themes that were originally deemed “unworthy of even attempting” back into a workable range—a scope where projects can be approved, prototypes can be built, and mistakes can be repeatedly made and corrected.
Why are these particular subjects popping up first?
This also explains a very intuitive phenomenon: The AI short dramas that initially appeared frequently were not office romance, family melodrama, nor were they the most cost-effective urban dialogue pieces, but rather genres like mythology, fantasy, adventure, and science fiction—elements that are inherently difficult to film practically.
The reasons are straightforward. For genres that are highly realistic, light on dramatic scenarios, and rely heavily on actors’ performance and the texture of life, the advantages AI currently offers are less direct. However, once a genre relies heavily on world-building, monsters, mythical creatures, large-scale set pieces, and non-realistic environments, the value of AI immediately becomes very tangible.
It first made the process of “imagining it out” much easier (or cheaper).
In content formats like short dramas—which are inherently highly sensitive to economic cycles, funding constraints, and risk tolerance—the party that lowers the cost of visualizing imagination first will be the first one to bring concepts and genres previously deemed unproducible onto the screen.
What Does Public Case Proof / Demonstrate
Based on public case studies, this line/trend has been very consistent.
On 2024-07-08, Kuaishou premiered The Legend of the Mountains and Seas’ Miraculous Mirror: Cleaving Waves and Cutting Waves, described as “China’s first original AIGC fantasy micro-short drama,” at the WAIC forum. This action itself is highly representative. What AI approached initially was not a low-cost reality genre, but rather a fantasy micro-short drama that requires strong world-building and visual effects support.
The AI short drama Journey to the West, launching on 2025-01-01, has switched its setting to Chinese mythology. Scenes like Flower Fruit Mountain, Water Curtain Cave, and Dragon Palace are inherently difficult to film practically with a small budget. What AI provides here is not just “the ability to generate images,” but rather the capability to create deliverable visuals for these scenes at a low cost first.
By 2025-02-24, The Mystery of Majia Cave’s Divine Staff Code shifted its theme to prehistoric tribes, archaeological fantasy, and totem narratives. A very crucial detail in the Science and Technology Daily report is that the team did not simply let AI generate random content; instead, they fed it research reports, 3D scans, and ancient texts, allowing it to make inferences within archaeological boundaries. In other words, AI is not merely cheap; it has transformed themes that previously required extremely high verification costs into an iterative production process.
Looking back, when CCTV.com reported on the premiere of “New World Loading” on 2025-06-26, it was no longer restricted to a single fantasy genre. Instead, it combined 7 unit dramas encompassing science fiction, fantasy, absurd comedy, history, and more. This signal is more important than any single case because it demonstrates that AI short dramas are moving beyond occasionally shooting supernatural mystery content; they are beginning to treat diverse genres, large worldviews, and style transitions as a continuously scalable production model.
What Appears to Be a Technical Gap Is Actually About Deliverability
Viewing only these cases, it is easy to misunderstand that “AI helps save everyone money on special effects.” The issue/scope is much wider than that.
By 2026-02-05, when Kuaishou released Kailin 3.0, the official focus was heavily placed on narrative control, consistency, videos of up to 15 seconds, and multi-character/multi-language native audio. When applied to short drama production, these terms translate into plain language as: Are the characters consistent (looking like the same person throughout)? Can the shots connect smoothly? Will the plot remain cohesive? And will scene transitions be jarring or structurally flawed?
This shows that AI short dramas are no longer just solving the problem of “whether it can be drawn first,” but are approaching “whether it can be connected like a finished deliverable product.”
On 2026-04-15, a Kuaishou executive publicly mentioned that Keling has reduced the cost of micro short dramas to less than one third of traditional models, and shortened the cycle by over 60%. I am more inclined to view this metric as an industry anchor point released by the platform side, rather than a general number applicable to the entire industry. However, it at least demonstrates one thing: that production methods are changing, and thematic boundaries will change accordingly.
When production shifts from “assessing resource sufficiency” to “evaluating conceptual potential,” those who will benefit first will certainly not be the most realistic genres, but rather those that require immense budgets—genres that are easiest to get killed by budget constraints.
Where are the remaining boundaries/limitations?
Of course, this does not mean that AI short dramas have replaced live-action human dramas.
It merely shifted the hurdle from mere feasibility (“whether it can be shot”) to narrative quality (“whether it is well-told”). Generating a spirit beast does not mean you automatically have a coherent Xuanhuan short drama; creating a war scene alone does not automatically establish character relationships, emotional pacing, or rhythmic control.
Many AI short dramas, even now, look more like “high-concept premise trailers” than complete narratives. They prove that the freedom to explore diverse themes has been released, not that their narrative competence has been adequately filled in.
So my judgment on this matter remains the same, though I’ve refined it since the previous draft: what AI did first for the short drama genre was not an aesthetic revolution, but rather putting a batch of previously scrapped concepts—those rejected due to perceived budget limitations—back onto the project proposal table.
Zombie, military, spirit beast fantasy are just the most obvious examples. There will be many more things later that shouldn’t appear in low-budget short dramas, which were brought into the test shooting area by AI. But what truly determines whether these genres can remain is still the script, pacing, and characters, not merely “finally being able to make it.”
References
Writing Notes
Original Prompt

AI short dramas have significantly expanded the themes of the short drama market: zombies, military, spiritual beast fantasy. Previously, live-action short dramas in these genres had production costs that were difficult to control. Using AI for this perfectly aligns with the needs of the industry.

Based on the original prompt above, this piece establishes its core theme, material density, and structure following the methodology of the first draft. The date field uses the original publication time, and all other content serves only the scope committed to by the current article.


What's truly terrifying isn't the layoffs, but the fact that they aren't hiring anymore.
Thu, 16 Apr 2026 02:18:44 +0800
Seeing Block cut its workforce by 4,000 people out of a group of over 10,000 at the end of February 2026 really shook me. I’ve always worked in financial IT—things like trading pipelines, Hong Kong/US stocks, and system fundamentals. Usually, I’m accustomed to buzzwords like “efficiency improvement,” “automation,” and “cost reduction while increasing efficiency.” But when a fintech company, one so close to money, compliance, and risk control, publicly cites AI as the reason for layoffs, it still hits you hard emotionally.
My current assessment is very direct: the scariest part of AI layoffs isn’t a layoff list in the news one day, but when companies start assuming that “smaller teams can do more work.” This means no backfilling for departures, fewer entry-level positions, and much tighter headcount management. From April 2025 to April 2026, this wind is still blowing in the US; while China hasn’t seen a high-profile, public wave of mass layoffs yet, the quiet squeeze has already begun.
Let’s look at these most representative companies first
Block was the harshest and most direct this time. After the Q1 2026 earnings report, Jack Dorsey’s statements left almost no room for ambiguity: AI has changed how the company builds and operates, allowing smaller teams to accomplish more. AP reported that Block laid off over 4,000 people, accounting for about 40% of its workforce of over 10,000 employees. The market was also honest; it rose by over 20% pre-market at one point. Why? Because investors were hearing not “technological ideals,” but profit margins, operating leverage, and faster earnings growth. Earlier on, Block had set targets during its Investor Day in November 2025 for a 17% increase in gross profit and over 30% growth in adjusted operating income for 2026. Simply put, this wasn’t about “laying people off because the company is failing,” but rather the company believing it could grow while simultaneously slimming down its workforce.
Duolingo, on the other hand, represents a different approach. At the end of April 2025, the company publicly shifted to an AI-first roadmap, stating that it would gradually phase out outsourced work that AI can handle, and any new hires must first prove that AI cannot do the job. A few days later, it launched 148 new courses created with generative AI. By August 2025, Duolingo’s earnings report showed a 41% year-over-year revenue increase in Q2, with strong profitability and subscription revenue, leading to a nearly 30% jump in stock price that day. However, the company also provided another lesson: users are not entirely convinced. The CEO later admitted that the backlash on social media indeed suppressed growth to the lower end of the guidance range. In other words, while AI can help a company scale its content production and boost profits, brand and trust are not freebies.
Klarna is perhaps the most typical example of “AI pressuring labor costs” over the past two years. It stated in 2024 that its AI customer service assistant handled two-thirds of inquiries, equivalent to the workload of 700 full-time customer service agents. In its prospectus submitted in March 2026, Klarna made the calculation even more explicit: at the end of 2023, there were 4,352 full-time employees; by the end of 2024, it was 3,422; and by the end of 2025, it would only have 2,831 people. The company itself stated that the reduction in headcount came from “using AI to improve efficiency and proactively compressing overall headcount.” Meanwhile, the average revenue per employee in 2025 was projected to rise from $344,000 in 2022 to $1.24 million. This figure is quite alarming. However, Klarna later began bringing some complex customer service issues back to human agents, especially those with high customer value and high risk scenarios. The lesson here is very clear: AI is excellent at absorbing standardized, repetitive, and process-driven tasks, but when it comes to complex judgment, trust endorsement, or gray-area responsibility, humans still need to be there as a fallback.
Shopify, however, has articulated something even scarier. In April 2025, Tobi Lutke’s memo to employees was very direct: before applying for more headcount, teams must first prove why the job cannot be done entirely with AI. It might not make daily headlines about layoffs, but it is creepier than a layoff list. Because it means that entry-level positions and roles that can be standardized and broken down are blocked at the source.
Why the Market Applauds
The capital markets love hearing AI stories; this is something that no longer needs to be hidden.
The reason is simple:

Revenue is not dropping, and is even growing.
Personnel costs will drop first.
Per capita output looks better.
The profit margin and cash flow potential in the future are much larger.

So, Block’s stock price will immediately rise because of the “smaller team can do more” narrative, Duolingo will receive good earnings feedback despite being criticized in the public eye, and Klarna will weave AI and efficiency into its IPO narrative.
Of course, there is real efficiency gain here, but there are also obvious elements of AI washing. Many companies aren’t seeing a scenario where some model suddenly becomes so smart that it can eliminate an entire department. Instead, management has finally found a nice-sounding reason to give to investors: to scale back the headcount expanded over the past few years, stop replacing people who naturally leave, and have AI work alongside fewer people on tasks like outsourcing, customer service, operations, content production, and low-level analysis.
How should I put it, AI is both a tool and a banner here.
Is this wind still in the US now?
It’s still there, and even more so by March 2026, the wind is stronger than in 2025.
Data from the US layoff tracking firm Challenger, Gray & Christmas speaks volumes:

In 2025, the number of layoffs citing AI as a reason reached 54,836 people.
In October 2025, AI was already the second largest cited reason for layoffs that month, corresponding to 31,039 positions.
By March 2026, AI became the top reason for layoffs cited by US companies for the first time that month, corresponding to 15,341 positions, accounting for about a quarter of the layoff plans disclosed that month.

But this dataset also needs to be looked at from another angle. It shows that AI layoffs are real, not just a joke; it also shows that we haven’t reached the point where “all layoffs are due to AI replacement.” Since Challenger started tracking this metric in 2023, those explicitly attributed to AI layoffs only account for a small fraction of all announced layoff plans. In other words, the US currently seems to be a combination of two forces:

One wave is genuine automation replacement, first hitting customer service, content, localization, junior white-collar workers, and back-end processes.
The other wave is macro environment, cost pressure, management’s narrative of contraction, topped with an AI veneer.

So, if you only look at news headlines, it’s easy to feel that “AI has eliminated white-collar jobs entirely”; but if you only look at official statements, it’s also easy to underestimate how it has begun rewriting hiring and organizational structures.
In China, why hasn’t it become like the US?
I tend to believe that it’s not that China hasn’t been affected, but rather that the way the impact manifests is different.
From the public narrative, China in 2025 to 2026 seems more like “scrambling for AI talent” rather than “publicly laying off people due to AI.”
According to a report by Xinhua News Agency at the end of March 2025, the job openings for algorithm engineers and machine learning positions in February saw year-on-year increases of 46.8% and 40.1%, respectively, indicating a significant supply shortage with an AI talent supply-demand ratio of approximately 3:1. By February 2026, another article from Xinhua News Agency mentioned that the hiring volume for AI-related positions in China in the fourth quarter of 2025 was still growing year-on-year by 19%. In terms disclosed by the Ministry of Industry and Information Technology in March 2026, the core AI industry scale in China in 2025 had already exceeded 1.2 trillion yuan, with related enterprises surpassing 6,200.
This set of data at least suggests two things:

China’s current public main themes are still industrial expansion, model implementation, and shortage of AI talent.
What is truly scarce is talent that combines algorithms, engineering, and industry, not ordinary white-collar office workers.

But this doesn’t mean there is no pressure.
Domestically, it’s more like a quiet squeeze:

Outsourcing is contracting first.
Hiring for campus recruitment and entry-level positions are more cautious.
Standardized roles such as Operations, Testing, Content, and Customer Service are being compressed first.
Some miscellaneous tasks that originally required two or three people to collaborate are now defaulting to one person working with AI.

So you might think, “Why haven’t I seen headlines about mass layoffs due to AI in China?” But the entry points for many positions have narrowed, they just haven’t been packaged as a public letter for investors.
Financial IT Development, What to Be Most Wary Of Isn’t Actually “Being Replaced Tomorrow”
I also work in financial IT, so I have a stronger feel for this part.
In the short term, development on core links within the financial industry—especially areas like trading, clearing, risk control, compliance, monitoring, and low-latency communication—are unlikely to be completely replaced by AI. The reason is simple: the stakes are too high, auditing is too strict, and the cost of an anomaly is too great; if something goes wrong, it’s not something you can fix just by changing a prompt.
But on the other hand, it is also very realistic:

People who write “glue code” will feel the pain first.
The value of pure CRUD, pure API lifting/moving, and pure documentation organization will continue to decline.
Tasks like testing, operations scripts, reporting, and internal tools will increasingly be assumed to be done with AI assistance (“you bring AI and finish it together”).
The requirements for a team’s individual combat capability, business understanding, and upstream/downstream connection ability will become higher.

To put it plainly, the most dangerous thing in the coming years might not be “AI replacing developers,” but rather that “a person who is proficient with AI and understands business boundaries will be considered equivalent to the output of two or three people from before.” If a developer only remains capable of rote coding tasks, without mastering the business domain, understanding risks, or taking ultimate responsibility, they will indeed become increasingly passive.
The Last, Less Pleasant Judgment
The news about Block shook me, not because it represents that AI is mature enough to completely take over fintech companies, but because it represents that management and the capital market are willing to say this, and the market is willing to believe it.
This is the most critical point.
Technology maturity, maybe not to that exaggerated extent yet; but organizations will first change based on “it’s almost good enough,” hiring will first change based on “it’s okay if we hire fewer people,” and promotions and performance reviews will also first change based on “why didn’t you use AI to do more.”
For average developers, especially in roles like financial IT that are caught between business needs, compliance, and system stability, what needs to be done next is not to fight AI head-on, nor is it to be blindly optimistic. Instead, it’s about quickly shifting oneself from being a “code laborer” toward becoming someone who “understands the business, understands the risks, knows the boundaries, and can provide fallback support.”
Otherwise, the wind from America will blow over eventually. Even if it’s not with the same news headlines, the pressure will fall on the same people.
References

AP News: Fintech company Block lays off 4,000 of its 10,000 staff, citing gains from AI
Block: Block Shares Multi-Year Financial Outlook at Investor Day
Block: Block Fourth Quarter 2025 Earnings Call
Klarna: AI assistant handles two-thirds of customer service chats in its first month
Klarna Group plc 20-F
TechCrunch: Klarna CEO says company will use humans to offer VIP customer service
TechCrunch: Duolingo launches 148 courses created with AI after sharing plans to replace contractors with AI
Duolingo: Q2 2025 results
TechCrunch: The backlash against Duolingo going ‘AI-first’ didn’t even matter
TechCrunch: Duolingo CEO says controversial AI memo was misunderstood
TechCrunch: Shopify CEO tells teams to consider using AI before growing headcount
Challenger: 2025 Year-End Challenger Report
Challenger: March 2026 Challenger Report
Xinhua News Agency: China’s rapid AI growth sparks hiring boom as demand outpaces supply
[Xinhua News Agency: AI reshapes China’s workforce via new professions, constant learning, one-person startups](https://english.news.cn/20260204/8d1bd05e8c4240998206fb9bdf

Writing Notes
Original Prompt

Compile the layoffs over the past year caused by AI, large-scale layoffs, the company’s operating status after the layoffs, market feedback, and whether the layoff trend is still present in the US or if it has little impact on mainland China. I myself am a developer in the financial IT industry, and the news of Block’s layoffs greatly shocked me.

Writing Outline Summary

Start with the Block news to maintain the personal perspective of “Financial IT developers are shaken.”
Instead of attributing all layoffs crudely to AI, break it down into three layers: “public layoffs,” “hiring freezes,” and “entry-level roles being squeezed.”
Select four representative examples—Block, Duolingo, Klarna, and Shopify—to examine layoffs, hiring freezes, operational results, and market feedback for each.
For the Chinese section, instead of stating it’s “nothing to worry about,” frame it as “the layoff wind isn’t as strong in China as in the US, but structural squeezing has already begun.”
Conclude by bringing the assessment back to the real situation of financial IT development; the focus should not be on emotional appeal, but on the changing professional structure.



Analyze the defaults and delinquencies in Chinese real estate.
Tue, 02 Dec 2025 19:06:37 +0800
Real estate giants’ top performers – Ever, also failed to hold up, continuous blood transfusions for Sunshine Steel, ultimately couldn’t plug the ever-growing hole. Linkage news: Six major banks completely halt sales of 5-year large savings bonds for five years.

Analyze China’s real estate defaults, compile timelines and key events, Evergrande, Greentown, Vanke, and other unlisted property developers I may not know about.


This overview is based on the latest perspective as of December 2025, covering the complete timeline from the initial crisis to the current period (end of 2025).
Currently, the landscape of China’s real estate industry can be summarized as: “Evergrande has liquidated, Hengda has just recovered, Greentown is still struggling, and Vanke is a golden body.”
Triggering Crisis: The Three Red Lines (2020)
The starting point was the regulatory authorities setting hard-line targets to curb real estate developers’ unrestrained expansion.

Time: August 2020
Policy: “Three Red Lines” (asset-liability ratio exceeding 70% after excluding prepaid receipts, net debt ratio exceeding 100%, and short-term cash ratio less than 1)
Consequences: Real estate developers could no longer use “borrowing and repaying” to snowball their debts, and the financial chain tightened instantly, leading to a domino effect of high-leverage real estate companies collapsing.

The Current Status and Timeline of the Big Three
Evergrande – The Catalyst of the Crisis

Current Status: Ordered into liquidation by the court. Xi Jinping has been subject to coercive measures, and the group has disintegrated.
Key Time Points:

December 2021: First formal default (unable to pay interest on dollar debt), designated as “restricted default” by rating agencies. This is a landmark event marking the beginning of China’s real estate crisis.
September 2023: Xi Jinping was taken into coercive measures for alleged illegal crimes.
January 2024: The High Court in Hong Kong formally issued a winding-up order, Evergrande’s overseas restructuring failed, and it entered liquidation proceedings.



Country Garden (“The Outstanding Student”) – The Fall

Current Situation: Debt restructuring in progress. As once the “No. 1 private real estate enterprise in the universe,” its defaults completely shattered the last confidence of the market in privately owned property companies.
Key Milestones:

August 2023: Acknowledged liquidity pressure and failed to pay interest on two dollar-denominated debt installments.
October 2023: Officially declared default (unable to pay a $470 million Hong Kong dollar installment) and hired advisors to initiate offshore debt restructuring.
2024-2025: Continued difficult debt restructuring negotiations and asset sales for self-rescue (such as Wanda Plaza equity, Australian projects, etc.).



Vanke – The Last Bastion (Latest Updates as of End of 2025)

Current Status: Breaking the “Rigidity”. Vanke had held up remarkably well for a long time, but by the end of 2025 it began seeking debt extensions, marking that mixed ownership/state-backed property giants couldn’t escape their fate either.
Key Time Points:

First Half of 2024: Faced short selling and credit rating downgrades, but thanks to verbal support from the Shenzhen Municipal Government and a syndicate loan (led by China Merchants Bank with a 20 billion yuan loan) it managed to pass through on time, repaying all publicly disclosed debt.
November 2025: First Proposal of Debt Extension. Vanke proposed extending payment for a soon-to-expire private placement debt financing tool (PPI), marking its first “technical default” or extension signal in public market debt, causing widespread shock in the market.



Other Notable Real Estate Giants with Defaulted/Breached Projects (Chronological Order)
In addition to the three you mentioned, numerous billion-dollar real estate companies have been involved. Here’s a list organized by “Default/Breach” time order:
Wave One: Early Meltdowns (2021)

China Fortune Land: Defaulted in early 2021, one of the earliest large property developers to collapse after the Three Red Lines policy. Currently has completed most of its debt restructuring.
Bluewray: Defaulted mid-2021, Sichuan’s “property tycoon,” delisted.
Fantasia: Defaulted in October 2021. Its default was extremely sudden and triggered extreme panic among the market regarding property developers’ “off-balance sheet debt.”
Kaisa: Secondary default in December 2021 (had defaulted in 2015). “The Old Renovation King” wasn’t spared either.
Sunshine City: Began defaulting at the end of 2021, a Fortune 500 company, rapidly collapsed.

Wave Two: Widespread Contagion (2022)

Sunac:

May 2022: Defaulted formally.
Status (2025): Offshore. Completed foreign debt restructuring in late 2023 and the restructuring plan officially took effect at the end of 2025. Sun Hongbin is one of the few bosses who successfully completed domestic and overseas debt restructuring, despite the company’s significant shrinkage in size, it preserved its operating entity.


Shimao Group: Defaulted in July 2022. Once known as the “Master of Luxury Homes,” it owned numerous landmark buildings (such as the Foshan Hotel) and is currently still struggling with debt.
CIFI Holding: Defaulted in October 2022. Formerly a “demonstration project” for private real estate companies, its collapse marked the complete annihilation of private property developers.
R&F Properties: One of the “Five Tigers of South China,” it underwent debt extension (although not formally termed as default) as early as 2022, and is currently selling overseas assets (such as London projects) to repay debts.

Wave Three: State-Owned/Mixed Ownership Model Under Pressure (2023-2025)

Sino-Ocean Group: Defaulted in 2023. This was the first state-owned property developer (backed by Ping An Insurance – a major shareholder) to default, shattering the “state-owned enterprise faith.”
Agile Property: Initially defaulted in the public market in May 2024 after a long period of not defaulting. Ultimately, it failed to hold on.

Summary: Real Estate Developer Survival Status Classification
To help you remember more clearly, we can categorize them into four classes:


“Dead”/Liquidation:

Evergrande (Liquidation), Logan Property (Delisted), Sunshine City (Delisted).



Currently “ICU” Restructuring:

Greenspun (In restructuring negotiations), SAMSING, Xinhai Group, Far East Land Development.



Restructuring Complete/Temporarily Stabilized (Moving to Ordinary Wards):

Rongcloud (Sun Hongbin self-rescue successful), China Wealth Securities.



Still Struggling/Just Showing Crisis:

Vanke (Just extending the term), Longhu, Jiuhua Group, Gold Land.




Related News: Regarding Vanke’s situation, I chatted with colleagues and friends in the group, and the conclusion was that interest rate cuts would continue. Shortly after, news flooded in about [Six Major Banks Fully Halt Sales of 5-Year Large Savings Bonds].  When reporters logged into the official apps and mobile banking platforms of the six major banks, they found that the maturity structure of large savings bonds had clearly become “short-term.” ICBC’s “Large Savings Bond” section only remained with products for terms of 1 month, 3 months, 6 months, 1 year, 2 years, and 3 years. Of these, the 3-year large savings bond product has a rate of 1.55%, and the 1-year and 2-year products have a rate of 1.20%. In addition, China Bank, Construction Bank, Bank of Communications, and Postal Savings Bank had similar product matrices, with all 5-year products removed from the sales list.  In Agricultural Bank’s RMB personal large savings bond products from 2018 to 2025, there was no trace of a 5-year large savings bond product. (China Finance News)


The End of the Age of Wild Capitalism: The Era of League of Legends Esports Concludes
Tue, 31 Dec 2024 07:19:39 +0800
The esports industry has experienced rapid development over the past decade and a half, becoming an increasingly significant cultural phenomenon worldwide. Particularly MOBA games like League of Legends (LoL), which is often abbreviated as LOL, not only provide players with the joy of competition but also inject powerful momentum into capital, driving the flourishing development of various esports platforms and events. However, all of this has entered a phase of gradual decline with the influx of capital and the rise of the entertainment industry. The rise and fall of PandyTV, the competition between DouYu and Huya, marks the end of the “wild capitalism era,” and the favorable conditions – time, place, people, and heaven – for esports seem to be starting to change.
Chapter 1: The Rise of Esports and Capital Injection
1.1 The Early Days of Electronic Sports: From Grassroots to Professionalization
The early esports industry started relatively grassroots, especially in China. Many players participated in competitive games like StarCraft and Dota, driven by their passion for the games. However, the true rise of esports began with the release and promotion of League of Legends. Following the official entry of League of Legends into the Chinese market in 2011, esports gradually evolved from a niche community to part of mainstream culture. Notably starting in 2013, the LPL (China Professional League) was established, and League of Legends became the cornerstone of China’s esports industry.
1.2 Capital Frenzy Influx: The Rise of PandaTV and Esports Live Streaming Platforms
2015 marked a watershed moment for the Chinese esports industry. The influx of capital transformed esports from a purely competitive event into a much larger ecosystem. PandaTV, one of the representatives, became a product of excessive capital. Invested by one of its founders, Wang Sicong, PandaTV quickly rose to prominence with its innovative content and massive investment, attracting a large number of viewers and users. However, this was also a typical example of capital “barbaric” inflow – the relentless pursuit of markets often lacks patience and long-term vision. Although PandaTV’s short-term investments achieved certain results due to management issues and excessive reliance on high fever caused by capital, it ultimately declared bankruptcy in 2019.
1.3 Live Streaming Platform Competition: The “Capital War” Between DouYu and Huya
The demise of PandaTV did not lead to the decline of the esports live streaming industry; instead, it propelled the rise of platforms like Douyu and Huya. Douyu and Huya became leading players in the esports industry, and their competition intensified. Early on, Douyu established itself as a benchmark for esports live streaming through the broadcasting of League of Legends professional competitions and the signing of top streamers. Huya, meanwhile, gradually narrowed the gap with Douyu by increasing its investment in esports events and diversifying its platform layout.
Throughout this process, capital played a huge role again. In 2018, Douyu successfully went public, and Huya also completed its IPO the same year. The rapid flow of capital led to industry consolidation and intensified competition between platforms in terms of streamers and copyrights, forming a “capital war.”
Chapter 2: The Fusion of Mass Entertainment and Esports
2.1 The Trend Towards Mass Entertainment: Capital Flows into Diversified Entertainment Projects
As capital has heavily invested in the esports industry, esports platform content is gradually becoming mass-entertained. Esports anchors are no longer limited to match commentary and event live streaming; they have begun to expand into singing, dancing, interactive live streams, and other entertainment forms. The content on platforms is becoming more diverse, gradually forming an ecosystem centered around esports but also incorporating various entertainment elements.
However, this mass entertainment trend has also brought problems – the original focus of esports culture has been marginalized, replaced by a trend of entertainment above all else. This trend has caused some fans who originally loved esports culture to feel alienated, and capital has begun to pay more attention to other entertainment fields. The excessive influx of capital and profit-seeking behavior have gradually blurred the essence of the esports industry, and the original value concept centered around competition has been weakened.
2.2 The Rise of Mass Entertainment Industry: Capital Withdrawal and Transformation
As short video platforms, live streaming platforms, and the entertainment industry have risen, capital has gradually shifted funds from esports to a wider range of entertainment content. In this process, giants such as Tencent, Alibaba, and ByteDance no longer rely solely on esports projects as sources of revenue, but are increasing their investments in areas such as film, music, and short videos.  In particular, the rise of ByteDance, through the explosive growth of platforms like Douyin (TikTok), has overshadowed the limelight of esports with other entertainment content.
Chapter 3: League of Legends’ “Yellow Not Following Yellow”: The Decline of the Era’s Advantage
Since League of Legends entered the Chinese market in 2011, it has almost become synonymous with China’s esports industry, achieving countless professional players, teams, and events, and also spawning a massive esports ecosystem. However, after more than ten years, as a leading project in Chinese esports, League of Legends has entered a phase of “Yellow Not Following Yellow” – a period of decline. Particularly in recent years, the influence of League of Legends is gradually waning, even exhibiting obvious signs of recession.
3.1 The “Fracture” in Player Groups
The most noticeable change is the fragmentation of player groups. Initially, esports’ rapid development relied on the support of a large number of young players, many of whom became professional players or spectators thanks to League of Legends. That generation of “internet addiction” teenagers almost grew up under the “era’s advantage,” immersed in the competitive charm of LOL and thus driving the rapid expansion of the entire industry. However, as time went on, these players gradually matured, entered society, and began to shift towards other life and career paths. At the same time, a new generation of young players are not as enthusiastic about League of Legends as in its heyday, and the esports audience has exhibited a clear age bias and decline in interest.
3.2 The “Weakness” in Game Content
League of Legends, after multiple updates and revisions, still maintains a certain competitive charm, but the game’s own innovation in content appears somewhat lacking. Each year’s version updates, hero balance adjustments, and new gameplay introductions seem unable to fundamentally address players’ demand for freshness. Meanwhile, the MOBA market has become saturated, and other types of games (such as Honor of Kings and Game For Peace) have rapidly risen, diverting a large number of players who originally belonged to League of Legends. This competitive situation has left League of Legends unable to escape the role of “follower.”
Conclusion: Where Does the Esports Industry Go From Here?
The esports industry is like a skyscraper that sprung up seemingly overnight, with excessive capital roaming the internet sector, searching for the next big trend – and esports has become one of their targets. Leveraging China’s demographic dividend, the industry achieved tremendous success in a short period, but this success wasn’t built on a solid foundation. Excessive capital influx, talent shortages, and weak game content are all hindering the healthy development of the esports industry.
Prior to university, I didn’t play many games; Alliance essentially accompanied my growth alongside generations. I watched countless World Finals, and as an outsider, comparing ourselves to Korean and Chinese players, especially Faker, domestic players always felt hesitant during major tournaments. I also know that athletes face immense psychological pressure, but this hasn’t been prioritized by teams. Over a decade of development, these mental health issues should be taken seriously, yet they weren’t. Domestic gameplay still relies on the innate talent of the players themselves.

AI Fish's Seven-Second News on Uncle Xiang's Notebook

Major AI Events

2022: The Gateway Opens

November: ChatGPT Goes Public (NVDA +25.4%)

December: From Demonstration to Diffusion (NVDA -13.6%)

2023: Capabilities, Open Weights, and Capital Expenditures Accelerate Simultaneously

January: Cloud and Office Begin to Bind (BBAI +385.2%)

February: Bard and LLaMA Fork (688256.SS +121.4%)

March: GPT-4 Begins Connecting to Tools (300308.SZ +52.8%)

April: Google DeepMind Merger (300476.SZ +52.8%)

May: Long Context Meets GPU Orders (SMCI +112.4%)

June: APIs Start Calling Functions (SOUN +49.2%)

July: Open Weights and Closed-Source Models Progress Together (SMCI +32.5%)

August: Code Llama Seizes the Code Entry Point (000158.SZ +26.2%)

September: ChatGPT Goes Multimodal (000158.SZ +21.9%)

October: Image Generation Enters the Dialogue (300502.SZ +47.5%)

November: DevDay and Governance Crisis (PLTR +35.5%)

December: Gemini Brings Multimodality into Products (BBAI +25.9%)

2024: Multimodality, Reasoning, and Agents Enter Products

January: GPT Store Attempts to Distribute Assistants (SMCI +86.3%)

February: Sora and the Long Context Race (SOUN +347.0%)

March: Models Begin to Stratify by Capability (300502.SZ +29.1%)

April: Llama 3 Expands the Open Ecosystem (300308.SZ -39.7%)

May: GPT-4o Enables Real-Time Interaction (NVDA +26.9%)

June: AI Enters Devices and Workflows (ARM +35.8%)

July: Large Models Shift Toward Low-Cost Routing (SOUN +28.9%)

August: AI Act Enters into Implementation (000158.SZ +69.3%)

September: A new curve emerges for reasoning models (000158.SZ +180.8%)

October: Agents begin operating computers and searching (SMCI -30.1%)

November: MCP Connects Tools and Data (SOUN +85.1%)

December: Sora Goes Live, Gemini Enters the Agent Era (SOUN +113.1%)

2025: Reasoning Costs Drop, Coding Becomes the Agent Proving Ground

January: R1 Recalculation Costs, Stargate Increases Infrastructure Investment (688158.SS +155.2%)

February: Research, reasoning, and coding extending task chains (300476.SZ +58.7%)

March: Web Search Becomes an API Capability (NVDA -13.2%)

April: Model Comparison Unit Becomes a Complete Task (PLTR +40.3%)

May: Claude 4 Competes for Long-Duration Coding Tasks (300308.SZ +56.5%)

June: Terminal Becomes an Agent Working Environment (601138.SS +65.1%)

July: Agent Moves from Model to Delivery (688256.SS +110.4%)

August: GPT-5 and Open Weights Advance Simultaneously (SOUN +26.0%)

September: Agent Needs Checkpointing and Rollback (MRVL +33.7%)

October: Low-latency model offloading (ARM +20.0%)

November: Gemini 3 Connected Development Platform (300502.SZ +23.9%)

December: Inference Capabilities Continue to Drive Down Costs (688158.SS +33.6%)

2026: Model companies begin to face scrutiny from the public market and real-world workflows

January: Constitution and Automatic Browsing (NVDA +2.5%)

February: Reasoning Models Move Toward Research Collaboration (2513.HK +20.6%)

March: Agent Completes Retrieval and Memory (688256.SS +72.9%)

April: Long Tasks Encounter Product Life Cycle (2513.HK +83.8%)

May: AI Applications Become Execution Systems (SMCI +68.2%)

June: On-device Models Meet the Open Market (MRVL +45.3%)

July: Coding Agent Becomes a Clear Main Theme (NVDA +5.4%)

What to Watch After the Timeline

The Entry Point Has Shifted from Chat to Workflow

Open Weights and Closed-Source Services Each Bear Different Costs

AIGC Moves from Content Generation to Credibility Competition

Coding is the earliest Agent scenario that can be accepted

AI Stocks Need to Move from Computing Power Orders to Cash Flow

Market Data and Record Boundaries

References

Original Prompt

Original Prompt of the First Draft

Raw Prompt for This Round

Trump Visits China Again: This China-US Leaders Meeting, First Seeking Stability, Then Discussing Transactions

Let’s first clarify this meeting.

The whole story was not actually sudden

First Paragraph Setup: The Busan meeting in October 2025 needs to slow things down.

Preamble for the Second Section: Leaders’ Call on February 4, 2026, putting this year’s agenda forward first

Third Section Preparation: The ROK Economic and Trade Consultation on May 12-13 Provides Background for the Beijing Talks

Contextual Background: The very timing of the visit to China itself also suggests that the situation is more pressing than what appears on the surface

To What Extent Has the Itinerary Been Disclosed?

Domestic View: First Steady, Then Discuss Differences, But the Red Line Will Not Loosen

International View: The spectacle will be large, the breakthroughs minor, but both sides cannot afford to miss this meeting

Viewpoint A: Strong emphasis on ritual and form, but limited substantial breakthroughs

Another View: Trump Needs This Meeting More Than He Did in 2017

Another layer of colder calculation: neither side wants to lose control

What exactly prompted this meeting?

1. The ceasefire following Busan requires extension

2. Both Parties Have Specific Transactions to Discuss

3. The Iran War Puts the US and China on the Same Table Again

Preamble for the Second Section: Leaders’ Call on `February 4, 2026`, putting this year’s agenda forward first

4. Both sides both require a relatively stable `2026`