ShizukaKazu

#Gate直通IPO第三期Anthropic The Claude experience suggests Anthropic is a company that builds models to be “colleagues who get things done”: they answer with tact and can directly help you write documents, edit spreadsheets, and work on projects.
· It talks to you like an adult: little fluff, clear logic. If your request is unclear, it asks you first.
· It handles long texts and code in depth: it can read an entire lengthy contract or codebase at once, then compare different sections. The code Claude Code writes is also more thorough.
· The chat window becomes a workspace: you can work on documents, slides, whiteboards, and spreadsheets right in the conversation, without copying content back and forth.
· It enforces safety strictly: there are fewer cases of mistakenly blocking harmless requests than before, but it still sometimes treats ordinary homework as a risk.
· There are trade-offs: usage limits on paid plans are rather tight, and complex tasks can easily use up your quota; registering and paying for an account in China is also a hassle.
So what Anthropic is really betting on is not just making models smarter, but enabling them to take on complete workflows like writing, research, and office work over the long term.

#Gate直通IPO第三期Anthropic Valuation shrinks by $600 billion, 7 founders prepare to cash out $250 million each—is Anthropic’s once-in-a-century IPO still attractive?
Anthropic is finally going public, but the price is a little hard to swallow. On November 10, Anthropic officially kicked off its IPO roadshow: 825 million shares at $70 to $78 each, raising up to $64.4 billion. In the proposal Anthropic sent to investment banks the night before, that figure was still $69.3 billion. Nearly $5 billion evaporated overnight. Anthropic’s own valuation also fell from the $2 trillion still being discussed by the market just two weeks ago to around $1.4 trillion, a drop of about 30%. That is already below SpaceX’s market cap of around $1.77 trillion at its IPO. More subtly, the founding team is cashing out even as the valuation is being cut.
According to the updated prospectus, 7 co-founders will sell existing shares in the IPO, cashing out a total of about $1.75 billion, or around $250 million each. Is this once-in-a-century IPO still attractive?
01 —Just how big is this once-in-a-century IPO?
First, the scale. In June, SpaceX set a global IPO record with a $75 billion offering, which came to around $85.7 billion including the overallotment. Anthropic’s public offering comes to around $64.4 billion at the top of the range; even if underwriters fully exercise the overallotment, it would only reach about $74 billion. Looking at the public offering alone, the record for the largest fundraising has not been broken. But it is still the second-largest IPO in history, about 2.5 times the size of Alibaba’s 2014 listing and 4 times Facebook’s 2012 listing.
What actually pushes the total raised past SpaceX is a concurrent private placement. This private placement totals $26 billion, subscribed to by seven tech giants at the IPO price. Google and Nvidia are putting in $10 billion each, while AMD, Broadcom, Micron, Microsoft, and TSMC split the remaining $6 billion. The public offering and private placement together bring the fundraising close to $100 billion—it really is money power at full throttle.
However, all seven private placement investors are upstream suppliers in Anthropic’s compute supply chain. Suppliers use their money to buy shares in their customer, and the customer then uses the money to buy compute from those suppliers. This model is called “circular trading” in the AI industry. The money seems to keep growing, but may simply be circulating within the supply chain. If end demand fails to keep up, both upstream and downstream players could be dragged down together.
So this IPO has slammed on the brakes amid a headlong rush. It looks glamorous on the surface, but there is no shortage of internal problems. Why has the capital market suddenly become so cautious? Anthropic’s operating performance is actually not bad; its revenue growth has been outrageous. /* QUOTE */ $386 million for all of 2024, $4.585 billion in 2025, and about $17 billion in a single quarter by the third quarter of this year. As of the end of October, annualized revenue was around $81 billion. Existing customers are also using more and more. Based on the prospectus, net revenue retention is about 900%. The same enterprise customers spend around 9 times as much on Claude a year later. Operating cash flow was also positive in the first half, at about $3.9 billion.
But Anthropic’s actual situation is not as rosy as its revenue curve suggests.
 First, growth is slowing.
Revenue grew 144% quarter over quarter in the second quarter, but that fell to 47% in the third quarter.
And Anthropic has already given back the money it just made. It posted its first operating profit of $974 million in the second quarter, then swung back to a loss of $567 million in the third. The main reason is that compute is too expensive and demand is growing too fast, forcing Anthropic to buy around $5 billion in short-term compute at high prices. Compute costs rose from 48% of revenue in the second quarter to 58%. On top of that, price cuts and contract discounts reduced third-quarter revenue by about $600 million.
Anthropic’s revenue also includes a less flattering source of income. According to the prospectus, in the first nine months, revenue from customers later found to have violated usage policies or commercial terms accounted for around 5% to 7% of total revenue, or about $1.7 billion to $2.3 billion. Most of this money came from users outside the US whose accounts were banned. Banning accounts without refunds has long been one of Anthropic’s go-to moves; who knew it was still a proper line of business on the books?
There is another counterintuitive point: the better Anthropic’s business does, the more it loses on paper. Its paper net loss for the first nine months reached a staggering $104 billion. More than $100 billion of that did not come from operating losses (the business itself lost only about $1.26 billion), but from accounting revaluations due to the increase in its valuation.
What does that mean? In its early days, some of Anthropic’s funding came in the form of “money now, shares later.” In other words, investors put in money upfront, with an agreement to convert it into company shares later under predetermined terms. These are the convertible notes and SAFEs mentioned in the prospectus.
The problem is that each quarter, the value of these “shares to be issued in the future” has to be recalculated using the latest valuation. In just over a year, Anthropic’s valuation rose from over $100 billion to more than $1 trillion. The value of those shares soared along with it, and the increase has to be recorded as a loss.
It’s like borrowing $100,000 from a friend to open a store, with an agreement that they can later convert it into a stake in the business. The more successful the store becomes, the more that stake is worth, and your books have to record it as a “loss.” That’s why the pressure from paper losses in Anthropic’s financial statements is so high.
On top of these pressures, Anthropic is also saddled with a compute bill of nearly $900 billion. This represents long-term purchase and lease commitments it signed with Google, Amazon, Broadcom, and others. Even if revenue stops growing, it still has to pay.
With such high spending, Anthropic has another, more fatal weakness: its customers and products are too concentrated. Its top five customers contributed about a quarter of revenue in the first nine months; more than 80% of third-quarter revenue came from models released in the past six months.
In other words, Anthropic has to keep running faster than everyone else. If it has a gap in new model releases, customers will leave in a heartbeat, and everything could be at stake in an instant.
02 —How will the share price move?
Before discussing price movements, let’s break down the key figures in this prospectus.
Anthropic plans to issue 825 million Class A shares, around 801 million of which are newly issued by the company. At the midpoint price of $74, this portion could bring the company about $59.3 billion. That money will be the fuel Anthropic needs for its upcoming compute battle. But the remaining 23.648 million shares are different. These are existing shares, and all of them will go into the pockets of the 7 founders.
Besides the share composition, another point worth watching in this offering is the post-listing share price stabilization mechanism. To prevent the share price from going on a “roller coaster” after listing, Anthropic has given the underwriters, led by Goldman Sachs and Morgan Stanley, a 30-day option to buy up to another 123.75 million shares at the offering price if necessary. This is called the greenshoe mechanism.
If the share price falls, the underwriters buy shares on the market to support it; if it rises, they get the company to issue the additional shares at the offering price. It’s like buying insurance for the share price: someone steps in if it falls, and the company can raise more money if it rises.
However, this insurance only lasts 30 days. SpaceX also used a greenshoe when it went public, but still fell below its offering price in mid-July.
So, how will Anthropic’s share price move after the 30-day insurance period ends?
First, the good news. Anthropic shares are in high demand right now. As soon as the roadshow began, orders were already fully subscribed within the price range, and institutional demand is even stronger than it was for SpaceX. But a big first-day jump is not guaranteed. First, the broader market is in bad shape. On November 5, the Nasdaq plunged 5.5% in a single day, its biggest drop in 19 months, with AI chip stocks tumbling across the board. Listing just as AI stocks are getting hammered is hardly ideal timing.
Second, SpaceX’s experience is still fresh. Its shares surged on the first day after listing in June, then fell below the offering price in July. Once the excitement fades, the market still looks at actual results.
There will also be a wave of selling pressure on the first day of trading. Under the prospectus, some directors, executives, and employees will be able to sell up to 10% of their holdings starting on the first day of listing.
Many Anthropic employees hold shares, most of which they received when the company’s valuation was still very low. Selling on listing day would let them make a big profit. If many people sell at once, the share price will be pushed down.
So, in the short term, sentiment matters; in the long term, it comes down to performance. The current offering price has already factored in a lot of expectations. The investment banks priced Anthropic based on its own revenue forecast for 2028: $190 billion to $200 billion.
As of the end of October, its annualized revenue was only about $81 billion. In other words, revenue needs to more than double over the next two years for today’s price to hold up. This IPO will affect more than just Anthropic itself.
Among the world’s leading large-model companies, Anthropic is the first to go public. Its share price is effectively setting a “benchmark” for the entire AI industry. If Anthropic rises, OpenAI’s valuation will have more support; if Anthropic falls below its offering price, the entire AI sector will be dragged down. Hong Kong-listed Zhipu and MiniMax won’t escape either.
03 Fable 5.2 drops the day before the roadshow
The day before the roadshow, Anthropic released its latest flagship model, Fable 5.2, just two months after the previous generation, Fable 5.1.
In terms of speed, it generates output up to 30% faster than 5.1; a new “fast mode” can speed it up by another 2.5 times.
In terms of price, input costs $8 per million tokens and output costs $40, reportedly 20% less than the previous generation.
In benchmark scores, Fable 5.2 scored 65 on the intelligence index from third-party firm Artificial Analysis, ranking first and ahead of Google Gemini 4 Carbon, which scored 60.
The most notable improvement is in coding. In the official FrontierCode 2.0 test, Fable 5.2 scored 49.1%, compared with just 32.4% for the previous generation, and also outperformed OpenAI’s GPT-6 Astra.
But benchmark scores are just on-paper results. How does it actually perform in practice?
Several viral hands-on tests on X gave it very positive ratings.
But as capable as Fable 5.2 is, Anthropic is actually deliberately holding it back. During training, Anthropic deliberately limited Fable 5.2’s capabilities in “accelerating AI development,” such as optimizing chip compute kernels and designing large-model architectures, keeping it at the level of Opus 5.5. Racing to build ever-better models while deliberately throttling its strongest model is the most contradictory thing about Anthropic.
04 Anthropic is rushing to list—why isn’t OpenAI in a hurry?
This year, three trillion-dollar companies are lining up to go public: SpaceX, Anthropic, and OpenAI. SpaceX went first in June, Anthropic followed close behind, while OpenAI chose to stay in the private market. On November 6, OpenAI completed the first closing of its latest funding round: $11 billion at a valuation of $1.25 trillion.
One is willing to go public even at a discount; the other would rather stay private than accept one. Why?
In fact, OpenAI has been preparing to go public for a long time. In June, a week after Anthropic, it also confidentially filed an IPO application. But in September, Altman said publicly that it would not go public in 2026. He cited AI safety: with so many safety issues right now, going public would be “unwise timing.”
But that may only be one reason. There are reports that advisers gave OpenAI two options long ago: wait until 2027 and go public at a trillion-dollar valuation, or lower its valuation and list as soon as possible. Altman rejected the latter. OpenAI’s CFO also privately recommended delaying, on the grounds that the company’s financial disclosures were not yet up to the standards of a public company.
Put simply, valuations in the private market are set through round after round of negotiations, without having to face scrutiny from shareholders every day. Once you go public, people score you every day, and every rise and fall is there for all to see.
So why does Anthropic insist on going public now? I think there are three main reasons.
First, it needs too much money.
As mentioned earlier, Anthropic is saddled with nearly $900 billion in compute commitments, while it has only about $132 billion in cash and short-term investments. Raising money round after round in the private market cannot keep up with its spending. This IPO and private placement will bring in nearly $100 billion at once.
Second, Anthropic’s books look good enough.
Going public means laying the books open for everyone to see. Anthropic is willing to go public because its revenue looks good, and it even briefly achieved operating profitability in the second quarter. OpenAI is under far more pressure: it expects to burn through about $278 billion in cash from 2026 to 2030.
Third, getting there first.
A huge pool of money around the world has been waiting for a “pure AI stock.”
The three companies are valued at more than $4 trillion combined, but there is only so much money in the market. Whoever lists first will be the first to capture that pool of capital.
05 Who is the real money-printing machine in the AI era?
Looking at the numbers alone, Anthropic is already ahead. As of the end of October, Anthropic’s annualized revenue was about $81 billion; the figure OpenAI gave investors was about $50 billion as of the end of September.
But these two figures cannot be compared directly, because the two companies account for revenue differently.
For example, if a customer spends $100 on Claude through a cloud platform, Anthropic records the full $100 as revenue and records the roughly $16 commission paid to the platform as a cost. OpenAI sells the same thing but only records the amount it receives after the commission. If OpenAI used Anthropic’s accounting method, the market estimates its figure at the end of September would be around $70 billion. Anthropic was about $71 billion at the same time.
In other words, the two companies’ revenue is actually about the same. The real difference is that their ways of making money are completely different.
However, OpenAI is catching up in the enterprise market. Since the third quarter, its enterprise revenue has more than doubled, now accounting for over half of its total.
Each model has its own weak points.
Anthropic’s money comes in fast, but it relies too heavily on big customers and new models. If its models fall behind, customers can walk away just like that.
OpenAI has many users, but most are free users. Supporting 1.2 billion people means compute alone costs an astronomical amount.
So, who is the money-printing machine?
Honestly, at this point, neither is. Nearly 60% of Anthropic’s revenue goes to compute costs, and the faster its revenue grows, the faster its compute bill grows; OpenAI still has several more years of spending to go. The companies collecting steady money are the ones selling them chips and cloud services. Google, Nvidia, Microsoft, TSMC… In this AI gold rush, it’s still the people selling shovels who make money first.
06 Conclusion
A company founded less than six years ago has reached a trillion-dollar valuation and made it onto the Nasdaq. AI is no longer just a technology in the lab, nor just a chat tool on your phone—it has become Wall Street’s most expensive business.
That is certainly exciting. But the faster money comes in, the more important it is to stay calm. A company’s valuation can swing by hundreds of billions of dollars in just a few weeks, showing that even the most professional investors don’t have a definitive answer.
This time, Anthropic may offer 15% to 20% of its shares to retail investors, far more than the 5% to 10% typical of an IPO. For many, this could be their first chance to buy shares in a leading AI company directly.
But being able to buy doesn’t mean it’s worth chasing. In the early days after a new stock lists, few shares are available to trade and sentiment is high, magnifying both gains and losses. If you chase the price up, you could easily get stuck holding the bag.
More importantly, you’re not buying a company that is already highly profitable; you’re buying into expectations for the next several years. The share price holds up only if those expectations are met; if they fall short, the price can drop just as ruthlessly.
So if you really want to buy, think through two things first: can you accept losing this money, and are you bullish on the AI industry or just the excitement around this stock’s first day of trading?
But one thing is certain. Over the past few years, we could only hear these companies tell their own stories about whether AI is a good business. After November 19, the market will answer that question every day.

#美国初请失业金人数高于预期 Initial jobless claims rose to 200k, still historically low; October core CPI, however, came in above expectations for the third consecutive month, and markets have shifted to pricing in rate hikes. That makes inflation the more important signal right now.
Why inflation signals matter more
The labor market is cooling only slightly: Initial jobless claims were 200k last week, the previous reading was revised to 191k, and continuing claims also came in below expectations.
Core inflation remains persistently hot: October core CPI rose 0.3% month over month, exceeding expectations for the third consecutive month.
Fed officials are hawkish: Several officials believe policy needs to be tightened further, and the probability of a 25-basis-point rate hike in December is about 84.7%.
Market outlook
Weaker employment and persistently high inflation point to more of a “recession-style rebound” for BTC and U.S. stocks than the start of a new rally. High interest rates weigh on valuations and add to deleveraging pressure.
BTC: Any rebound could quickly be capped by a stronger dollar and high Treasury yields, with price action more volatile than U.S. stocks.
U.S. stocks: AI-related tech stocks have support, but valuations are vulnerable to interest-rate shocks; sectors such as consumer-related industries and healthcare, with stable cash flows, are relatively resilient.
Gold: Near-term support is weak as hopes for rate cuts fade.
Trading approach
It is better to wait and see for now, until the December Fed meeting outcome and new inflation data are available. If your positions are large, consider reducing leverage and keeping ample cash rather than rushing to buy the dip. If employment deteriorates significantly for four consecutive weeks while core CPI falls, then consider increasing risk exposure. A one-week rise in initial jobless claims to 200k is not enough to change this outlook.#WCTCS9

#BTC跌4% Bitcoin Falls Below 71,000: It’s Not Just the Price That’s Falling—The Market Has Lost Its “Common Anchor”
It fell 1.79% yesterday and 4.12% today. The drop is accelerating, and that in itself is more worth watching than the exact level it has fallen to.
But focusing only on price means missing something more important. Grayscale published a report on November 12 that made a key point: over the past three years, Bitcoin’s price movements explained only about 50% of the return differences across crypto market sectors.
Put another way, Bitcoin is no longer the market’s single, common anchor. This shift happens to explain why so many people are confused right now: Why do some still feel they “can’t make sense of things” even though the Fear and Greed Index is at 20? Because sentiment indicators are still unified, while pricing logic has already fractured.
01 Market action: The drop widened from 1.79% to 4.12%
First, the decline widened from 1.79% to 4.12%, while the Fear and Greed Index fell only from 28 to 20. If sentiment is a function of price, that decline is in line; but if sentiment is a function of expectations, it suggests concerns about what comes next weigh more heavily than the drop that has already happened.
Second, the intraday low of 70,688 has already reached the edge just above 70,000. This level happens to be an area that was repeatedly discussed during the previous market cycle. When prices enter this kind of zone, the focus shifts from “fundamentals” to “positioning”—because further downside triggers not valuation judgments, but stop-loss orders.
02 This Grayscale report matters more than the price
The report’s core consists of three sets of figures, each worth unpacking.
First: Bitcoin’s price movements explain only about 50% of return differences across crypto market sectors. This means the other half comes from factors specific to the sectors and individual assets—for example, a sector’s narrative, a project’s progress, or the liquidity structure of a particular asset class. In other words, the “watch BTC to trade every coin” approach now has only half the explanatory power it used to.
Second: Correlations between crypto sectors have fallen from previous highs, and market performance is becoming increasingly divergent. This is a direct result of the first point. When correlations are high, the market behaves like a single whole, rising and falling in sync; when correlations fall, the market becomes several submarkets, each priced on its own.
Third: Annualized volatility varies by sector—about 43% for Bitcoin, 49% for payments, and as high as 85% for AI.
These figures are the most informative: they show that “diversification” changes not the absolute level of risk, but its structure.
Grayscale itself makes this point in the report: lower correlations do not mean lower investment risk.
Taken together, these three sets of figures point to one conclusion: the crypto market is shifting from “one beta market” to “multiple alpha markets.” In the past, you only needed to assess Bitcoin’s direction; now you also have to judge which sectors capital is flowing into and how they are being priced. This hasn’t made things simpler—it’s made them harder. And that helps explain why the current market action looks like it is “falling for no reason.” When the market still has a common anchor, declines have a unified explanation; once that anchor comes loose, declines become “each falling on its own,” which makes them look like they have “no reason.”
03 Fear and Greed Index at 20: Sentiment is still unified, but logic has fractured
There’s a contradiction here that’s easy to overlook. The Fear and Greed Index has fallen from 28 to 20, entering “extreme fear.” This is a unified sentiment indicator—it doesn’t distinguish between sectors or assets, but gives the entire market a single score.
But Grayscale tells us that pricing logic has already diverged. This creates a mismatch: using a unified sentiment indicator to explain a market that has already fractured. The result is that the index tells you “everyone is fearful,” but some sectors in your portfolio have fallen much more than BTC, while others have fallen less. You can’t use “fear” as a single variable to explain them all.
The more practical issue is that a reading of extreme fear cuts both ways. On the one hand, it means selling pressure has been heavily released, and the most pessimistic moments often coincide with short-term bottoming zones; on the other hand, fear can persist for a long time, and “extreme fear” followed by more “extreme fear” is the norm. So this index can tell you the “position,” but not the “timing.” That’s also why this piece doesn’t say whether you “should buy the dip.” In a market that has lost its common anchor, the term “buy the dip” is itself imprecise—you first have to specify which sector’s bottom you’re buying.
04 So what is the nature of this decline?
Putting the first three sections together, we can offer a more accurate description than “the bad news is all priced in.” This decline wasn’t caused by any one piece of bad news, but by the combination of “the disappearance of the pricing anchor” and “persistent selling pressure.”
The previous day, we had already seen consecutive net outflows from ETFs and arbitrage traders unwinding basis trades—that was structural buying stepping aside; today, the decline doubled and the Fear and Greed Index shifted within a single day—that was sentiment catching up on the downside. When the two coincide, prices take on a “the further they fall, the faster they drop” shape.
And the disappearance of the “common anchor” changes how the decline spreads.
It used to be “BTC falls → the whole market falls”; now it’s “BTC falls → sectors fall in order according to their own liquidity.”
Sectors with poorer liquidity and narratives more dependent on sentiment fall earlier and harder; sectors with genuine independent drivers may even move in the opposite direction.
That’s the other side of what Grayscale calls “diversified sources of returns”: it also means diversified sources of losses. Lower correlations don’t reduce risk; they just turn “losing together” into “losing separately.”
05 Several signals worth watching next
No predictions about direction—just things that can be verified. One Whether the decline narrows.
A move from -1.79% to -4.12% is an acceleration; if the decline narrows noticeably on a given trading day and no new low is made, that’s the first sign that selling pressure is easing
Two Whether the Fear and Greed Index finds a floor around 20.
If it continues falling toward single digits, sentiment is still deteriorating; if it starts to recover, that usually corresponds to short-term selling pressure having run its course
Three Whether sector declines start to diverge.
By Grayscale’s figures, AI sector volatility is twice that of BTC—if AI-related assets fall significantly more than BTC during the decline, that’s a normal “queuing by liquidity” transmission; if the reverse happens, it means capital is rotating between sectors
Four Whether ETF flows shift from persistent net outflows to narrowing outflows.
This is direct evidence of whether structural buying is returning
Five Price action around 70,000.
Watch just one thing: after a breakdown, does the price quickly recover, or does it move lower on heavy volume? The former is a liquidity gap; the latter is a trend
The easiest thing to overlook in this downturn: while the market is still looking for a “common anchor,” the most common mistake is using Bitcoin’s price action to explain every asset. Grayscale’s 50% figure is a reminder of this habit—it doesn’t say Bitcoin is unimportant; it says “Bitcoin is no longer enough.”$BTC ‌#WCTCS9

#ETH一度跌破2000美元 ETH falling below $2,000 doesn’t mean it has bottomed. Reclaiming $2,000 would only signal that the decline has stopped; a rebound would only be underway once ETF inflows return and the price holds steady.
Why it’s still too early to call a bottom
· Buyers haven’t returned: On November 5, spot ETH ETFs saw net outflows of about $430 million, with none of the 12 products recording net inflows.
· Leverage hasn’t been fully flushed out: Long positions accounted for 90% of liquidations, but funding rates are still positive, indicating that bulls are still paying to hold their positions.
· The technical picture remains weak: The 50-day moving average is overhead, and the rebound has been capped near $2,050, with no reversal signal forming.
What to watch for to confirm a move back above $2,000
· Net inflows to ETFs for several consecutive days: A genuine turnaround in flows.
· A close above $2,058: A breakout on rising volume, with moving averages starting to flatten.
· Funding rates turning negative: Bulls are no longer dominant, and liquidation risk is falling.
· BTC stabilizing at the same time: ETH is unlikely to establish a trend on its own.
What I would do
I wouldn’t rush to buy the dip or sell everything. If heavily positioned, I’d reduce leverage first; if looking to add, I’d wait for several consecutive days of ETF net inflows and for ETH to close and hold above $2,058, then enter in stages. The $2,000 round-number level alone isn’t enough to determine whether the bottom is in.$ETH ‌#WCTCS9

#Dr.Han牛津辩论社演讲 Combining the views expressed by Dr. Han in his “Start Alone” speech at the Oxford Union, I believe the core capabilities entrepreneurs need most in the AI era can be summed up in three points: the ability to think strategically about using AI, the resilience to weather cycles, and the organizational ability to connect people. AI boosts efficiency, while people set the direction, shoulder the pressure, and bring people together.
I. The ability to think strategically about using AI — the key to “Boss+AI”
• Dr. Han’s “Boss+AI” model is essentially about making AI an amplifier, not a substitute for thinking.
• Entrepreneurs should delegate information gathering, learning, and the efficiency of day-to-day work to AI, while keeping decisions about direction and values in their own hands.
• The key is knowing how to use AI and how to ask: understanding what AI can and cannot solve, and being able to break vague ideas into actionable instructions.
• People who know how to ask good questions will move faster in the future than those who only know how to pile up tools.
II. The resilience to weather cycles — the foundation of “Start Alone”
• Dr. Han founded Gate on his own without external venture capital and weathered industry fluctuations and business challenges through long-term conviction.
• In the AI era, opportunities come and go quickly, and trends shift frequently. Entrepreneurs who lack resilience can easily be shaken out.
• Resilience is not brute force; it is the ability to “experiment and adjust quickly”—to reflect after failure and persevere through downturns.
• Rapid innovation is important, but it must be grounded in a stable core; otherwise, it is just blind pursuit.
III. The organizational ability to connect people — the other side of “Start Alone”
• Dr. Han specifically emphasized that “Start Alone” does not mean going it alone forever. Entrepreneurs need to find the right partners and team as they grow.
• AI can replace repetitive labor, but it cannot replace trust, collaboration, or a shared vision.
• Entrepreneurs need to be good at recognizing, deploying, and retaining talent, and uniting the team so they can truly realize AI’s efficiency advantages.
To sum up: In the AI era, the technical barrier is falling and the tools are available to everyone. The real dividing line is whether you understand the direction better than AI, can withstand more than your competitors, and are better than machines at bringing people together. These are the core capabilities AI can never provide for you.

#Dr.Han牛津辩论社演讲 Combining the latest speech by Dr.Han at the Oxford Union, the answer is actually very clear: sound judgment is absolutely central, innovation is the differentiating moat, and AI has already made execution efficiency an industry standard. In the AI era, solo entrepreneurship means “the barrier to getting started has dropped significantly, but the bar for long-term survival is actually higher.”
🎯 First priority: sound judgment is the difference between life and death
In his speech, Dr.Han specifically mentioned that Gate began after he encountered setbacks trading in his early years and resolved to build a more trustworthy platform. He had no outside venture capital backing at any point and weathered more than a dozen rounds of industry volatility on his own.
• AI can help you research an entire industry and generate hundreds of proposals, but which sector to enter, when to make your move, and whether you can withstand market cycles are ultimately decisions only you can make.
• If you choose the wrong direction, even AI’s powerful execution efficiency will only help you move faster down the wrong path—and fail sooner.
💡 Second priority: innovation is the moat
Now that everyone can use AI, basic productivity tools are no longer a secret. Anything you can do with AI, your competitors can do too.
• The “Boss+AI” model Dr.Han proposed is essentially about breaking out of the habit of “using AI to do menial tasks” and turning AI into an amplifier for your innovation, rather than merely an execution tool.
• What truly sets you apart is whether you can use AI to combine things in new ways and create new experiences and business models that others haven’t seen. That’s what others can’t copy.
⚡ Last priority: execution efficiency is the baseline
This is precisely what AI excels at most: writing copy, designing, coding, and processing data—AI can do in minutes what used to take a team a week.
• Execution efficiency has now become a “baseline requirement” for entrepreneurship, not a “bonus.”
• You don’t need to push yourself to the limit in this area; AI has already raised the floor remarkably high.
🚀 Solo entrepreneurship: the barrier has dropped, but the challenge has changed
Starting a business on your own in the AI era really is much easier than it was 10 years ago:
• In the past, you needed to bring together several people for technology, operations, and marketing to launch a project. Now, one person can use AI to run the entire process.
• In the past, you had to spend months doing research and building a product. Now, you can produce a minimum viable product in just a few weeks and quickly test your idea.
But it hasn’t become “easier”:
• In the past, you had to shoulder the pressure of “having no one to do the work.” Now, you have to contend with intense competition as “hundreds of new rivals use AI to enter the market every day.”
• As Dr.Han emphasized in his speech, “Start Alone” doesn’t mean going it alone forever. You can get started on your own and use AI to make things work, but to truly grow into a platform like Gate that serves more than 60 million users, you’ll eventually need to learn how to build a team.
Simply put, AI has lowered the “difficulty of getting started” with solo entrepreneurship, but raised the bar for “decision quality” and “differentiation through innovation.”

#黄仁勋辞职 #每周来晒 Jensen Huang’s resignation changes Nvidia’s long-term valuation rationale, but demand for AI computing power itself has not changed. The sharp stock decline is mainly the market removing the “Jensen premium.”
For Nvidia
· Stock price: Closed down 4.34% on November 12, after falling about 14% at one point intraday; its market cap remains above $4 trillion.
· Valuation reset: Goldman Sachs cut its price target from $300 to $220, while Bernstein lowered its target to $200.
· Orders remain: Institutions have generally not lowered earnings expectations, only compressed the price-to-earnings multiple.
For AI-related stocks
· Nvidia supply chain: Stocks such as optical module makers and suppliers will face short-term pressure.
· Alternatives: AMD, Google TPUs, and domestic AI chips will benefit, while cloud providers will also accelerate procurement diversification.
· Third-party computing power: Computing-power companies relying on financing to expand capacity are the most vulnerable, as financing may tighten.
Whether demand for AI computing power is weakening is mainly judged by Nvidia’s orders and earnings.
The new CEO has not yet been decided; stock stabilization will have to wait for earnings and clarity on the new CEO.
Domestic AI chip companies are the biggest beneficiaries among A-shares.
Two things to watch next: the November 17 earnings report, to see whether orders have shifted elsewhere; and whether the new CEO can stabilize customers and the supply chain. $NVDA ‌

#AnthropicIPO六倍超额认购 Anthropic will see a divergent performance after going public, with a “rapid short-term surge followed by sharp mid-term volatility,” rather than sustained one-way strength.
🚀 In the early days after listing: a certain surge
1. Sixfold oversubscription was driven entirely by large institutional orders. Underwriters are already considering pushing the offering price above the $70–$78 range, with the valuation potentially soaring to $1.5 trillion.
2. It will likely be added to six major mainstream indexes within 5–15 days of listing, triggering nearly $20 billion in mandatory passive fund buying. Its performance will directly replicate this year’s SpaceX surge playbook.
3. Retail investors’ net buying on the first day is expected to reach $250 million, double SpaceX’s previous record, further fueling short-term sentiment.
⚠️ After the mid-term: high valuation will inevitably bring sharp volatility
1. Its current $1.4 trillion valuation essentially discounts $685 billion in revenue in 2030 in advance, implying 150-fold growth over five years—a feat without precedent in business history.
2. It won’t achieve positive free cash flow until 2030. Annual computing capacity expenditures of tens of billions of dollars are a long-term, heavy burden, and the $90 billion fundraising will only provide short-term relief.
3. Shares allocated strategically to the seven major tech giants have no lockup restrictions, so large-scale selling could emerge at any time once the stock surges.
Simply put, it could surge sharply in its first half-month after listing, buoyed by passive index funds and sentiment. But after the rally, the huge gap between its high valuation and fundamentals will inevitably drag the stock into wide fluctuations. #WCTCS9

#BTC跌4% Jensen Huang resigns, so why is the crypto market plunging across the board? Bitcoin briefly fell below $68,000
The man who turned graphics cards into strategic assets and made “computing power is power” the era’s gospel is gone.
Jensen Huang has resigned as president, CEO, and board director, effective immediately. CFO Colette Kress has been appointed interim CEO in a crisis, while semiconductor veteran Harvey C. Jones becomes the first chairman of NVIDIA’s board.
After the announcement, NVIDIA shares briefly plunged 13% in premarket trading, and its intraday market cap fell below $4 trillion for the first time since July last year. By the 13th, the shockwave had crossed Silicon Valley and hit the crypto market squarely—Bitcoin fell below $69,000, Ethereum lost the $2,000 level, and SOL, XRP, and BNB all bled. Coinglass data showed $1.232 billion in liquidations across the market in 24 hours, with more than 150,000 traders liquidated.
Why did the resignation of a graphics-card seller’s boss leave the crypto market with a liquidity shortfall?
01 First, clarify the facts: not a rumor, but an official handover
1 Time: NVIDIA’s board announced it on November 12 (U.S. Eastern Time); the SEC disclosed it simultaneously.
2 Move: Huang stepped down as president, CEO, and director, exiting management entirely.
3 Succession: Colette Kress is interim CEO and CFO; Harvey C. Jones is the first chairman.4 Search: A formal search for a permanent CEO is underway, both internally and externally; no decision has been made.
5 Background: The board had previously launched an independent investigation into disputes over Huang’s private life and governance. In four days, things went from “cooperating with the investigation” to a “complete exit.”
This is not an “extended vacation” or a “move to chairman.” It is a hard break with strongman rule.
02 Why was the crypto market “detonated from afar”?
On the surface, NVIDIA and Bitcoin have nothing to do with each other. Look deeper, and the 2026 crypto market is no longer an independent market—it’s a shadow market for AI risk appetite.
1. Narrative link: The AI standard-bearer falls, and risk assets are repriced across the board
For the past two years, “AI stocks + crypto assets” have been the same kind of position in institutional portfolios—a bet on technological revolution, liquidity spillover, and future discounting.
Huang is no ordinary CEO; he is the personification of “AI certainty.” When he steps down, the market sees not a personnel change, but NVIDIA moving from the “Jensen Huang premium” into a “professional-manager discount” phase. Demand for AI computing power hasn’t disappeared, and the CUDA ecosystem hasn’t collapsed, but with the narrative leader changing hands, valuation anchors are already wobbling. When tech stocks shiver, the crypto market catches a cold.
2. Leverage cascade: Longs get “liquidated with one click”
The worst losses in this downturn have been in derivatives, not spot markets. About $1.23 billion was liquidated in 24 hours (some estimates put it at $720 million, due to different reporting windows); BTC longs made up the vast majority, with the largest single liquidation at $31.09 million (Hyperliquid); three eight-figure long positions were liquidated in succession within five minutes. The largest single liquidation was $31.09 million; three eight-figure liquidations in five minutes. A ton of people were using leverage to go long at $73,000–$74,000, waiting for a rebound after “the NVIDIA bad news was fully priced in.”
Instead, the bad news wasn’t fully priced in—it turned into a funeral. Stop-loss orders triggered → prices broke lower → more stop-losses triggered: a classic liquidity meat grinder.
3. AI tokens’ “knee-jerk reaction”
The AI sector, most closely tied to NVIDIA’s narrative, took the hardest hit: AI tokens such as NEAR, Render, and Bittensor led the decline; the entire AI token sector fell 5%–10% in 24 hours.
The logic is simple: NVIDIA is the pricing center for AI assets. When the center changes hands, speculative assets on the periphery are the first to be repriced lower.
4. The macro powder keg was already in place
Huang was just the match. The barrel was filled with:
U.S. CPI above expectationsOctober CPI came in higher than expected, reviving expectations of a December rate hike.
Dollar and Treasuries riseThe dollar index surged above 103, while Treasury yields also climbed.
Stablecoin supply shrinks by $14.2 billionSupply contracted by $14.2 billion over 30 days (Glassnode figures, the fastest pace since 2022).
ETF outflows continueSpot Bitcoin ETFs continued to see outflows, led by BlackRock’s IBIT.
In a market without liquidity, bad news gets amplified threefold.
03 A personal take: What this downturn really means
More precisely, NVIDIA’s leadership upheaval was the “final emotional blow” that crushed leveraged positions in crypto; but the underlying cause was crypto’s own liquidity crunch + a synchronized weakening of risk assets.
First, crypto fell harder than NVIDIA because it’s more “intangible”
NVIDIA still has data center orders, margins, and the CUDA moat; the crypto market has no single earnings anchor and survives purely on liquidity and narrative. When the narrative wobbles, it’s the first to fall.
Second, “AI + crypto” moving in sync is a double-edged sword
In a bull market, AI lifts crypto; in a bear market, trouble in AI drags crypto down too. In 2026, the two can no longer be separated.
Third, this isn’t a bubble bursting; it’s leverage being flushed out
There’s currently no stablecoin depeg, no exchange run, and no LUNA-style structural collapse.
It looks more like this: the highly leveraged longs built up in the first half of the year are being washed out in a single day.
04 What to watch next
Don’t rush to declare “the bull market is dead,” and don’t buy the dip just yet.
Watch three signals:
Signal one: NVIDIA’s November 17 earningsData center growth and customer capex guidance will determine whether valuations can hold up in the “post-Huang era.” If they can’t, tech stocks will keep falling, and crypto won’t escape unscathed.
Signal two: Whether stablecoin supply stops falling
If the $14.2 billion contraction continues, it means outside capital is still pulling out. Without ammunition, any rebound is a fake move.
Signal three: BTC support in the $68,000–$71,000 range
If this level holds, it means a sideways base is forming after leverage has been flushed out; if it breaks, the next liquidity pool is at $62,000–$64,000.
In 33 years, Jensen Huang turned NVIDIA into the “central bank” of the AI era. The way he exited wasn’t dignified, but it was shocking enough—the market has finally realized that even the strongest narrative can’t withstand the founder leaving, and even the fiercest bull market can’t survive excessive leverage. What took a fall in this crypto plunge wasn’t Bitcoin, but the belief that it would “go up forever.” In the post-Huang era, the first thing to be liquidated isn’t NVIDIA’s valuation—it’s the crypto market’s naivety. #WCTCS9 $BTC ‌

#黄仁勋辞职 #每周来晒 33 Years, $4 Trillion, An Overnight Curtain Call: The End of the Jensen Huang Era—Where Does Nvidia Go from Here?
01 Breaking: First CEO Change in 33 Years
On November 12, U.S. Eastern Time, Nvidia published an announcement on its website: co-founder Jensen Huang resigned as president, chief executive officer, and board member, effective immediately.
The announcement gave no reason and provided no transition period. This is the first CEO change since Nvidia was founded in 1993. In the days before, reports concerning his personal affairs continued to gain traction, and Nvidia’s board had launched an independent investigation; details remain subject to company announcements and subsequent disclosures by authoritative media. From the news breaking to his formal resignation, only four days passed.
More revealing than the resignation itself is the transition plan announced alongside it.
Colette Kress | Interim CEO
The current executive vice president and chief financial officer joined Nvidia in September 2013 and has experienced the company’s entire journey from graphics chips to a leading AI computing powerhouse. Until a permanent successor is appointed, she will also continue to serve as finance chief.
Harvey C. Jones | Board Chair
The former CEO of Synopsys becomes the first independent board chair in Nvidia’s history. The company had never had such a position in its previous 33 years.
CEO Search | Internal and External Candidates
The board has launched a formal search, with both internal and external candidates under consideration. No timeline has been given.
Bloomberg said the board has engaged an external executive search firm. The words “effective immediately” sever 33 years of continuity. No transition period, no deputy, no publicly named successor—that is what the market truly sees.
02 From $40k to $4 Trillion
To understand the significance of this resignation, we must first look back at Jensen Huang’s journey.
Thirty-three years, one company, one person—there are almost no comparable examples.
1993 | A Silicon Valley Restaurant
At 30, Jensen Huang founded Nvidia with Chris Malachowsky and Curtis Priem, with $40k in startup capital. Back then, the word “GPU” did not yet exist, and gaming graphics cards were a niche market.
1999 | GeForce 256
Nvidia launched what is widely recognized as the world’s first GPU, transforming the graphics processor from an accessory to the CPU into an independent computing unit.
2006 | CUDA
The most visionary and most questioned bet of his career: insisting that GPUs should do work beyond graphics. Wall Street once thought this was a waste of money.
2012 | AlexNet
A deep neural network trained on two GTX 580s won ImageNet. The deep learning era began, and GPUs became the computing foundation of AI.
2022–2026 | From Graphics Card Maker to “Picks-and-Shovels Seller”
ChatGPT triggered global demand for computing power, and H100s became impossible to get; Nvidia’s market cap successively surpassed $3 trillion and $4 trillion, making it one of the most valuable companies in human history.
It all began with an architecture sketched on a napkin and $40k. This is one of the very few Silicon Valley founders who “founded the company and took the helm, then steered it to the summit.”
Beyond Musk and Zuckerberg, it is hard to name a third.
03 What Did the “Leather-Jacketed Tough Guy” Leave Behind?
Jensen Huang left the tech world far more than just a company.
A management philosophy
He dislikes hierarchical barriers and departmental silos, and does not favor one-on-one meetings; instead, he convenes management as a group. At the peak, 60 direct reports reported to him; as of this November, the number was still 33.
One thing he often says is: “Our company is always just thirty days away from going out of business.”
A long-term mindset
He did not give up on CUDA after more than a decade of continuous losses; even in the darkest moments, when crypto was in decline and the gaming business was falling, he still bet everything on AI.
A “picks-and-shovels seller” positioning
He does not build large language models or applications, only the foundational computing power and software ecosystem that everyone depends on. This made him a “scarce resource” rather than a “competitor.”
A very small equity stake
He holds only about 871 million common shares, representing 3.58%, yet controls a $4 trillion company. His control does not come from his shares, but from the personal trust he has built over 33 years—and that is precisely the hardest part to hand over.
04 Without Jensen Huang, Can Nvidia Keep Printing Money?
The market has already delivered its answer!
After the announcement, Nvidia’s U.S. premarket losses quickly widened to about 10%; it briefly plunged about 14% after the open, and closed down 4.3% at $166.19. Compared with the October 6 high of $239.24, the stock has now pulled back about 30%; its market cap shed more than $600 billion in a week, equivalent to about 1.2 trillion yuan. On the same day, AMD and Broadcom each rose more than 5%, while Cerebras surged after hours. The bet is not on Nvidia’s collapse, but on orders being spread across suppliers. On November 17, the company will release its earnings as scheduled—that will be the new management team’s first time facing the market alone.
More worth watching than the stock price is the web beyond the balance sheet.
Over the past two years, Nvidia has turned “investment equals customer” into a growth engine through purchase commitments, project guarantees, and equity investments: providing up to $105 billion in phased, conditional guarantees for OpenAI’s data center in Ohio; committing to buy up to $36 billion in cloud services from “neocloud” providers that purchase its chips; and providing up to $125 billion in support for a $500 billion credit program.
Columbia Business School professor Stijn Van Nieuwerburgh estimates that these guarantees and commitments total about $400 billion. Hence two descriptions: some call Nvidia the “central bank of AI,” while others ask whether it could become the Lehman Brothers of the AI era.
Former FDIC Chair Sheila Bair’s assessment is measured and clear: Nvidia is not a bank and will not face a bank run; demand for its core chips remains, and its massive equity holdings provide a thick cushion. But if it really collapsed, it would be an earthquake-level event.
The real issue is that Nvidia has never had a clear succession plan.
Jensen Huang has dozens of direct reports, major decisions are concentrated in one person, and project financing documents even have to wait until he lands before they can be signed off.
Bernstein’s analysts put it most bluntly: “Jensen Huang is Nvidia, and Nvidia is Jensen Huang.”
The odds on potential successors have already appeared on prediction markets. In contracts launched on November 9, the four leading candidates are:
Ian Buck | 32%
He joined Nvidia in 2004, was one of the core developers of the CUDA platform, and is currently vice president of accelerated computing and high-performance computing, working directly with Microsoft, Google, Amazon, Meta, and OpenAI.
Colette Kress | 22%
The CFO of 13 years, she has appeared alongside Jensen Huang at every earnings call. When the market needs a steady voice, she is the most familiar one.
Rene Haas | 9%
The CEO of Arm, he previously worked at Nvidia for seven years and is familiar with the semiconductor industry—a “logical” external option.
Ajay Puri | 3%
One of the company’s longest-serving core executives, with 21 years at Nvidia, he long oversaw global sales and customer relations and became a senior adviser in August this year.
A person spending 33 years at one company, taking it from a corner of a restaurant to the pinnacle of global market capitalization, is a miracle in itself in business history. But the business world never stops for anyone.
The question Nvidia must answer next is: Is the company “Jensen Huang’s Nvidia,” or a machine that can already run itself? Perhaps we will find out at the next product launch—but the person on stage will no longer be wearing that leather jacket.

#Gate直通IPO第三期Anthropic Anthropic’s prospects are essentially those of a “superstar at commercializing AI,” but half of its future is in its own hands, while the other half is hitched to the global compute industry’s wagon.
🚀 Core growth drivers (proven with certainty)
• Explosive revenue growth: Revenue was just $41 million in 2023, surged to $4.6 billion in 2025, and reached $16.3 billion in the first half of 2026 alone. Annualized revenue has already approached $81 billion, with a $10 billion net increase in a single month. This growth rate is an order of magnitude faster than traditional SaaS.
• Enterprise willingness to pay is off the charts: Net revenue retention is as high as 900%, meaning that existing customers’ annual spending can grow to 9x the baseline. What starts as a small pilot quickly becomes core AI infrastructure across the company. It now has over 1,000 enterprise customers spending more than $1 million a year.
• Model capabilities continue to lead: The newly released Claude Fable 5.2 has left GPT-6 Astra and Gemini 4 Carbon behind in benchmarks across core use cases such as intelligent coding, office automation, and scientific research simulations. Its edge is especially overwhelming in high-value areas like robotic simulation and cyber offense and defense.
• Unique positioning across all cloud ecosystems: It is the only frontier model available on all three top cloud platforms—AWS, Google Cloud, and Microsoft Azure—giving it direct access to over 90% of global enterprise customers without having to build a sales network from scratch.
💰 Capital markets’ sky-high expectations
Lead underwriter Goldman Sachs has made a striking prediction: Anthropic’s revenue will reach $685 billion by 2030, including $586 billion from enterprises and the remaining $100 billion from consumer subscriptions. That would be 150 times its 2025 revenue.
Its current $1.4 trillion IPO valuation implies an annualized price-to-sales ratio of 17.3x. Though at the very top of industry valuations, Anthropic has become the global AI sector’s valuation benchmark—its IPO performance will directly shape risk appetite for capital across the compute supply chain and the large-model sector.
⚠️ The existential tests hanging over it
• Compute commitments are its biggest burden: It has already signed nearly $900 billion in long-term compute procurement commitments. By 2030, compute spending is set to rise to nearly $400 billion, meaning it will have to burn enormous sums every year on chips and data center rentals. Goldman Sachs estimates it won’t achieve positive full-year free cash flow until 2030.
• Growth cannot stay frenzied forever: Its own prospectus explicitly warns that the current hypergrowth is unsustainable. As its revenue base grows, the inflection point from “winning new customers” to “digging deeper into profits” will arrive soon.
• The double-edged sword of cooperation and competition with giants: Google, Amazon, and NVIDIA are its cornerstone investors and compute providers, while also being direct competitors in the large-model sector. This “funding it while competing for its business” relationship could change at any time.
🎯 Final assessment
Anthropic has secured a top-tier ticket to the AI era. It is not a bubble propped up by storytelling; it has already demonstrated the commercial value of large models through real enterprise spending.
But the road ahead is not about continuing to burn cash on compute to drive growth. It must turn today’s “high growth, high losses” into a virtuous cycle of “high profits, high barriers to entry.” If it can weather the peak of compute capex over the next 3–5 years, it has every chance of becoming a tech giant on the scale of Apple and Microsoft today.$ANTHROPIC ‌

#直通IPO第三期Anthropic $69.3 Billion IPO: AI Bubble or the Next Decade’s Major Bull Run?
Anthropic’s IPO roadshow may be the most dramatic moment in the 2026 capital markets.
The day before, the market was discussing an offering plan seeking to raise about $69.3 billion at a valuation approaching $2 trillion; the next day, the filing was withdrawn and replaced with a more restrained offering range of $70–$78 per share, with the public offering’s fundraising cap adjusted to about $64.4 billion. Combined with a $26 billion private placement, the total could reach about $90.4 billion. Valuation expectations also fell from $2 trillion to $1.1 trillion–$1.4 trillion. This was not simply a pricing mistake, but a market repricing of AI assets.
Anthropic’s prospectus is half myth, half bill.
The myth is dazzling: as of the end of October 2026, annualized revenue was about $81 billion; third-quarter revenue was $16.96 billion, up more than 11-fold year over year; and the number of enterprise customers paying more than $100,000 annually reached 11,900. From $41 million in 2023 to $81 billion today, this growth curve has almost outpaced Google’s and Meta’s pre-IPO growth.
But the bill is just as staggering: the company has signed commitments to purchase about $900.9 billion in computing capacity and infrastructure, with about $70.7 billion due in the next 12 months. In the third quarter, the company’s operating profit turned into a $567 million loss from a $974 million profit, as it temporarily procured computing capacity at higher prices.
The structure of the participants is more complex still: seven giants—Google, Nvidia, AMD, Broadcom, Micron, Microsoft, and TSMC—have subscribed to a combined $26 billion in private placements. They are customers and shareholders of Anthropic, as well as suppliers of computing capacity and chips. This creates a closed loop: Anthropic trades equity for computing capacity, computing suppliers trade orders for equity, and companies across the industry chain boost one another.
Is this a bubble or a new cycle? More accurately, it may be both.
The bubble side is that valuations are still based on the assumption of “continued ultra-high growth in the future.”
If AI model prices keep falling, open-source models catch up, and corporate budgets tighten, trillion-dollar valuations will look burdensome. Especially as sequential revenue growth falls from 144% to 47%, the market has already begun recalculating its margin of safety.
The new-cycle side is that AI is no longer just a concept; it is entering enterprise workflows.
Claude Code, Claude API, enterprise knowledge bases, and automated workflows are moving from “trials” to “paid use.” When a company’s annualized revenue exceeds $80 billion, it can no longer be simply classified as a “story stock.”
The real question is not whether AI has value, but where that value will accrue.
Over the past two years, capital markets rewarded “more powerful models”; going forward, the market will care more about “who can make more money.” Can computing costs come down? Can customer retention be sustained? Can the application layer generate stable cash flow? These are the factors that will determine whether AI enters a major bull run or faces a valuation shakeout.
For ordinary investors, a more practical approach is to treat it as a barometer of the AI cycle: if Anthropic ultimately prices its shares conservatively and the overallotment option is exercised, that would indicate continued confidence in leading AI assets, and sectors such as computing capacity, chips, and cloud services could continue to benefit; if the offering price is lowered and subscriptions are tepid, it could mean the AI sector is entering a period of stress testing, making it more dangerous to chase concept stocks; if the stock fluctuates sharply after listing, ordinary investors should be even more careful to avoid emotional buying, especially bets made with leverage or concentrated in a single asset.
For most investors, rather than chasing the halo of the “largest IPO in history,” it is better to focus on AI applications that can genuinely generate cash flow, reduce costs, and improve efficiency.
The next wave of AI opportunities may not lie in the model companies with the highest valuations, but in the application layer that turns AI into a productivity tool.
Anthropic’s IPO is not just a pricing experiment for the capital markets, but also a watershed moment as the AI industry moves from “telling stories” to “reporting earnings.” Trillion-dollar valuations can be exciting, but what is truly worth watching is always cash flow, customer spending, and the path to profitability. The trend will not disappear, but the people most likely to lose money in it are often those who see only the valuation and not the bill.
👇What do you think? Welcome to share your thoughts in the comments～$ANTHROPIC ‌

#Gate全球合规版图扩展 Compliance capabilities are becoming as critical a barrier as liquidity.
Why compliance is becoming increasingly important
Institutional funds only enter licensed platforms: without U.S. licenses and auditable custody, compliance departments won't approve them.
Offshore arbitrage opportunities are shrinking: the EU's MiCA is now fully in effect, and the Middle East and Southeast Asia are also tightening regulations.
Smaller platforms are being squeezed out first: AscendEX, for example, halted services over licensing and compliance issues.
How to view Gate's footprint
Broad coverage: 81 jurisdictions spanning crypto trading, custody, payments, and other areas.
Its strength lies in securing high-value licenses in several key markets, with separate licenses in each jurisdiction. This carries more weight than simply having a large number of licenses.
High-value licenses: It holds Malta MiCA and Dubai VARA licenses, as well as Japan FSA and Australia AUSTRAC registrations.
Separate licenses in each region: In Europe, it has both a Lithuania registration and a Malta MiCA license, providing greater resilience to regulatory risks.
Broad coverage of U.S. state licenses: Gate US has obtained about 30 state money transmitter licenses, covering 40 jurisdictions.
High reserve transparency: Its total reserve ratio is about 127%, disclosed using zero-knowledge proofs; RootData ranks it second globally for transparency.
These licenses can help Gate enter more markets and make it easier to attract institutional funds.

#美国10月CPI全面超预期 #每周来晒 The Warning Sign of Rising Rents—Commentary on October U.S. CPI
October U.S. core CPI exceeded expectations. Particular attention should be paid to the rise in the rent component: it has bottomed out and rebounded, which could make a return of inflation to the 2% target seem indefinitely out of reach. In recent months, month-over-month increases in rent and housing services have continued to rise, showing signs of breaking higher. In addition, U.S. rents, a leading indicator, had already begun rising earlier. Given their high weight, this means inflationary pressure could surge in 2027. If the relevant components show no improvement in the November-December data, the 2027 rate-hike path may need to be revised sharply higher. On the one hand, a higher inflation baseline will require more tightening by the Fed; on the other, the current rise in rents may itself be a negative feedback response to excessively high mortgage costs (people are not buying homes but renting), which rate hikes cannot effectively curb. This means prices of other consumer goods will have to be brought down further, requiring additional Fed rate hikes. For markets, the endpoint of the rate-hike cycle will be difficult to determine in the near term, which is unfavorable for all asset classes.
October core CPI rose broadly again, showing signs of intensifying inflationary pressure
In October, U.S. core CPI rose 0.30% month over month, versus 0.33% last month. Although it edged down, the absolute level remains high, and the annualized rate is far above the target. Among the major components, month-over-month increases in core services were significantly higher overall than in core goods, and most components posted relatively high monthly increases, rather than the rise being driven by any single component. Compared with September, growth accelerated in medical equipment and auto insurance, while used cars declined markedly.
Structurally, watch the significant risk of a sustained rise in the rent component
Last month, core CPI unexpectedly rose to 0.33% month over month, with housing services increasing 0.4% month over month, making the largest contribution. This month, housing services eased slightly month over month but remained at 0.3%, with the owners’ equivalent rent component flat at 0.3% month over month. This development warrants close attention: the bottoming and rebound in the rent component could make a return of inflation to the 2% target seem indefinitely out of reach.
Housing services account for more than 30% of overall CPI. Their weight in PCE is somewhat lower, but still in the 15-20% range, and their trajectory has a significant impact on overall inflation. During 2024-2025, one major reason overall core CPI was able to decline steadily was that rent-related CPI indicators entered a long-term downcycle, with month-over-month increases in housing services remaining in the low range of 0.1-0.2%. Now, however, housing services have risen by more than 0.25% month over month for three consecutive months and by more than 0.3% for two consecutive months, seemingly indicating a higher baseline. In addition, rent, a core component of housing services, typically lags real-world rents in the CPI by about one year. Since 2026, U.S. rents have begun a new upswing (the market believes this may be related to high home prices, heavy mortgage burdens, and housing demand shifting toward the rental market). If this trend is extrapolated linearly, the CPI rent component will also face sustained upward pressure in 2027. Once month-over-month growth in housing services overall remains at 0.3% or even higher, given its substantial weight, it will become extremely difficult for the core CPI month-over-month baseline to return to 0.2%. That would require greater suppression of other components, potentially raising the bar for Fed rate hikes.
Policy implications and market impact
Pressure on the Fed to raise rates will increase further. Watch the November-December data; if the rent and housing services components do not improve, the 2027 rate-hike path may need to be revised sharply higher. On the one hand, as noted above, if rent- and housing services-related components push the core CPI month-over-month baseline back above 0.3%, inflationary pressure will rise, requiring more tightening by the Fed; on the other, the current rise in rents may itself be a negative feedback response to excessively high mortgage costs, which rate hikes cannot effectively curb. This means prices of other consumer goods will have to be brought down further, requiring additional Fed rate hikes.
For markets, the endpoint of the rate-hike cycle will be difficult to determine in the near term, which is unfavorable for all asset classes. The drag from high interest rates will persist, making valuation expansion difficult, while U.S. equities remain highly volatile; an opportunity for a sustained decline in Treasury yields will have to wait, and gold’s safe-haven appeal will also be diminished.

#Gate股票暢聊季 CoreWeave does have debt problems; they just haven't affected its growth story yet
CoreWeave is indeed highly profitable and has very strong business growth, but looking only at revenue makes it easy to overlook the most dangerous part of its business model. CoreWeave expects full-year revenue to reach $13 billion to $13.3 billion, exceeding Wall Street's previous expectations. The market sees this as a sign that demand for AI computing power remains strong and customers are still willing to pay for GPUs. But revenue growth does not mean improved cash flow, much less stronger debt repayment capacity.
According to the company's disclosures, operating cash flow for the first three quarters of 2026 was about $5.64B, while spending on property, plant and equipment and capitalized software reached $23.56B. The difference leaves a free cash flow shortfall of about $17.92 billion.
What does this mean? For every dollar of operating cash flow CoreWeave earns, it has to spend even more on buying GPUs, building data centers, and expanding computing infrastructure. It must keep raising financing to sustain its business expansion.
More concerning, the company expects interest expense in the fourth quarter alone to reach $1.05 billion to $1.15 billion. At that rate, interest expense could exceed $4 billion over the next year.
Its financing moves look especially interesting in this light. After reporting its results, the company's share price surged, but after hours it announced plans to raise at least $4 billion through highly rated loans.
Although the company's own credit rating remains junk grade, it can use chips and related contracts as collateral to obtain relatively cheap financing.
There is a financial structure here that is well worth studying. GPUs are both CoreWeave's most important means of production and important collateral for obtaining financing. Lenders are looking not only at the company's profitability, but also at these expensive chips and the computing contracts behind them.
In other words, AI computing assets are becoming a new financing tool. It sounds clever, but that's also where the risk lies. If GPU prices keep rising, demand for computing leases keeps growing, and customers pay on time, this structure can certainly work. But if AI capital spending suddenly slows, GPU rental prices fall, and customers start cutting orders, the value of the collateral, contract cash flows, and financing capacity could all come under pressure at once.
This reminds me of the subprime lending structure before the 2008 financial crisis. Of course, the two cannot be directly equated, but their financial logic is similar: rising asset prices support collateral values, the collateral supports more debt, and the debt drives further asset expansion. As long as growth expectations hold, everyone feels safe.
The real problem is that once growth expectations change, risks that previously seemed independent can erupt all at once.
That's why I've never liked simply defining CoreWeave as a beneficiary of AI infrastructure. It's more like a quintessential example of financialization during an AI boom: the business growth is real, customer demand is real, but so is its dependence on continued financing.
The most dangerous thing isn't that a company has debt; it's when the market mistakes financing capacity for profitability, and revenue growth for a mature business model.$CRWV ‌

#GateIPOAccessAnthropic The grey market is already up 33%, and the golden buying opportunity closed five weeks ago
This IPO is neither a golden buying opportunity nor a chance to escape. The grey market is already up 33%, and the golden buying opportunity closed five weeks ago! This is “a high-valuation growth stock offering that was priced in ahead of time.” The line between a bargain and an opportunity isn’t the company—it’s the price.
I. Look at the market, not just the prospectus
The prospectus says: 825 million shares, $70 to $78, a fully diluted valuation of $1.26 trillion to $1.41 trillion, pricing on November 18, trading on the 19th, and ticker ANT.
But the market is already ahead of it: subscription orders have reached several times the shares on offer, underwriters are considering pushing the valuation above $1.5 trillion, and the final offer price could exceed the $78 upper limit.
The grey market makes this even clearer. The xyz:ANT contract, which settles against the IPO price, is already trading at $103.91, about 33% above the $78 upper limit, implying a diluted valuation of about $1.91 trillion. Some institutions that received allocations early are already transferring existing shares through SPVs.
There’s another bit of context: five weeks ago, the valuation high being discussed by the market was $2 trillion; now it’s been cut straight to $1.4 trillion. This “halving of expectations” is itself a textbook case of expectation management.
II. How to tell “bargain or escape”
First, is there still a discount?
The purpose of the pricing range is to give primary-market buyers a margin of safety. The grey market is now already 33% above the upper limit, so that margin of safety is gone. If you get an allocation, you’re buying at the “offer price,” but the grey-market price is the reference for the first day of trading—the gap has already been pocketed by someone else.
Second, who’s selling?
Of the 825 million shares this time, the company itself is issuing about 801 million, while existing shareholders are selling only about 23.65 million; the seven founders are also cashing out about $1.8 billion in total. Against total proceeds of $90.4 billion, the cash-out accounts for only around 2%.
So this isn’t a “mass escape.” Most of the money is going into the company’s coffers to fund computing capacity—it now has close to $900 billion in computing expenditure commitments. This is additional funding for expansion, not shareholders selling out and leaving.
Third, is there still a case for calling the valuation cheap?
At a fully diluted valuation of $1.4 trillion and annualized third-quarter revenue of about $68 billion, the price-to-sales ratio is 20.7x; compared with OpenAI’s $1.25 trillion valuation on $50 billion in annualized revenue, or 25x, ANT is about 20% cheaper.
But at the grey-market valuation of $1.91 trillion, the price-to-sales ratio is 28x, making it more expensive than OpenAI.
So the “relatively cheap” argument holds only within the $70 to $78 range. Once the price moves higher, the discount disappears. That’s why I think the final price will most likely land at $80 to $85—any higher, and applying for shares is no longer worthwhile.
III. Three players, three different answers
For the company: this is a chance to raise funds—$90.4 billion, surpassing SpaceX to become the largest public listing in history, with the money going toward building computing capacity.
For the founders and early shareholders: this is a chance to exit and cash out—with $1.8 billion cashed out and existing shares transferred through SPVs, selling in this round at a valuation of $1.9 trillion is already a very good deal.
For retail investors: this is a lottery ticket with low allocation odds, a possible increase in the offer price, and a first-day premium already reflected in the grey market. The risk-reward is worse than it was five weeks ago, and perhaps even worse than waiting for a pullback after the listing.
IV. If you must participate, here’s what I’d do
First, price. Participate only in the lower part of the range ($70 to $73); if the final price exceeds $85, I’ll walk away—that would put the price-to-sales ratio above 24x, making it less attractive than buying from its shareholders.
Second, position size. No more than 5% of total assets; treat it as an event-driven trade, not a long-term holding.
Third, timing. If it rises more than 30% on the first day, sell half; if it falls 10% below the offer price, don’t add; around May 2027 is when the six-month lockup expires, and I’d reduce the position to zero before then.
V. The real risk isn’t that it’s “expensive”
Third-quarter revenue was $16.956 billion, up about 12x year over year and 47% quarter over quarter; in the same period, the GAAP net loss was $23.3 billion. The company says this includes large non-cash items related to preferred shares—so you can’t simply say it “burned $23.3 billion in cash.” But the close to $900 billion in computing commitments is real.
This means the company will have to keep raising funds over the coming years. If capital markets tighten up on AI, it will become “the most expensive capital-intensive company in the market.” This isn’t a valuation issue; it’s a cash flow issue.
So I’ll be watching two numbers: whether gross margin can hold in the mid-60% range (API is currently above 80%), and whether revenue can maintain quarter-over-quarter growth above 30%. A decline in either of these numbers would be the real signal.
Would you apply for shares, or wait for it to fall 30% after listing before considering it? Tell me in the comments.

#WCTCS9 This rebound doesn't look particularly reliable. Tech stocks are rising on CoreWeave's better-than-expected results, while Bitcoin is mostly benefiting from short covering. Neither has confirmed a reversal.
AI tech stocks: Earnings support, but macro conditions are weak
Main driver of the rebound: CoreWeave's third-quarter revenue grew 168% year over year, and it also raised its full-year guidance.
A-shares riding on optical modules: On November 11, the ChiNext Index rose 3.11%, but excluding three stocks including Zhongji Innolight, almost all of the gains disappeared.
Rate hikes weighing on valuations: Core CPI rose 0.3% month over month in October, and the probability of a 25-basis-point rate hike in December is about 91%.
BTC: Still hasn't caught up
Price remains low: Around $73,731 on the morning of November 12, up just 0.15% in 24 hours.
Gains driven by short covering: Short liquidations accounted for over 70% on the 11th, while trading volume actually fell.
ETFs failed to catch the inflows: U.S. spot ETFs saw about $1.28 billion in outflows for the week ending November 6.
My trading strategy: Wait for three signals
1. Bitcoin moves back above $77,600 and stays there for three consecutive days.
2. Spot ETFs post net inflows for several consecutive days, rather than reversing after a single day.
3. The December rate decision is announced, with the hike no larger than expected. For now, smaller positions are more suitable for short-term trades. If you want to add to your position, it's safer to wait until all three signals appear.