Googles new model is good enough that even if
1) OpenAI went bankrupt
2) Anthropic went bankrupt
3) Chinese open source was banned
For most things it actually wouldn’t matter
Big leap and I think will be a brutal overhang on already decelerating inference market
The oil price has NEVER dropped this fast unless the economy was in a complete crash..
Is that really what we are amidst? It certainly doesn't feel like it..
This tells me that this is supply driven rather than demand driven
The problem is the assets of the wealthy are mostly made up of the debts of the people whose jobs AI will take according to👇.
As those jobs are lost, those debts will default & those assets will collapse, which will then take down banking systems & sovereign bond markets. 1/
A Generational Mistake in the Making
The Fed is preparing to make a historic, generational mistake.
The real policy rate is already 1.93%, well into restrictive territory. Now the base effects are about to tighten it further, with no help from the Fed. Next spring, the war’s energy spike drops out of the inflation numbers.
Leave policy where it is and the real rate climbs to 2.21% by the end of 2027 even if energy prices only hold flat. If energy returns to its pre-war level, it reaches 3.29%, the highest since 2000. If the energy glut pushes prices 10% below pre-war, inflation falls under 1% and the real rate hits 3.88%, the highest since 1990.
That’s the tightening coming from doing nothing. The Fed wants to hike on top of it.
Core goods inflation is 0.7%. Input prices in the New York Fed’s own survey have plunged. Oil and diesel are falling. Yet the case for hiking rests on anecdotes about AI and a fear of growth, with no evidence offered.
Arthur Burns let inflation run in the 1970s. This Fed is making the opposite mistake: hiking into falling inflation and sending real rates to levels not seen in a generation. History will not be kind.
Intuit bought Mailchimp for $12B in 2021 and just sold to Bending Spoons for $3.6B, reflecting a -21% IRR on the investment
Meanwhile, revenues have grown 59% under their ownership, from $800M to $1.3B
The entire decline in the purchase price is from the multiple going from 15x to 3x
But sure, your SaaS private equity firms who are showing the same marks on their PortCos as 3 years ago are safe!
Stanley Druckenmiller:
"So, I’ll never forget it. January of 2000 I go into Soros’s office and I say I’m selling all the tech stocks, selling everything. This is crazy…at 104 times earnings. This is nuts.
Just kind of as I explained earlier, we’re going to step aside, wait for the next fat pitch. I didn’t fire the two gunslingers. They didn’t have enough money to really hurt the fund, but they started making 3 percent a day and I’m out. It is driving me nuts.
I mean their little account is like up 50 percent on the year. I think Quantum was up seven. It’s just sitting there. So like around March I could feel it coming. I just, I had to play. I couldn’t help myself. And three times the same week I pick up a, don’t do it. Don’t do it.
Anyway, I pick up the phone finally. I think I missed the top by an hour. I bought $6 billion worth of tech stocks, and in six weeks I had left Soros and I had lost $3 billion in that one play.
You asked me what I learned. I didn’t learn anything. I already knew that I wasn’t supposed to do that. I was just an emotional basket case and couldn’t help myself. So, maybe I learned not to do it again, but I already knew that."
Huang Statement (excerpt):
"I did not live up to the standards expected of everyone at Nvidia. I created problems for the Board that I deeply regret. I own my mistakes completely.
Since my departure, I’ve seen an outpouring of calls from Nvidia’s employees, investors, business partners, and customers for me to return as CEO.
If the Board of Directors decides that my return is in the best interest of our shareholders, our employees, the company, and America, I would be honored to answer these calls and return as CEO."
Reasons not to hike rates at all, let alone 50bps:
1. Oil just crashed 20%
2. CPI is backward-looking
3. Inflation is at the average of the last 40 years
4. Stocks crashed last week
5. It's one print. An ugly print, but one print, and the Fed shouldn't be making decisions based on one data point
6. Inflation was a lot higher in 2021 and there was no urgency then, why, I wonder, does everyone have religion about this all of a sudden?
7. You never hike into a supply shock
8. Everyone knows this is about politics
9. Everyone knows Warsh is strongly disliked
10. And Warsh is a proxy for Trump
11. Copper just broke the uptrend
12. A 50bp move is for emergencies. Nothing here constitutes an emergency.
13. See my previous tweet
Jensen was probably too naive.
Look at how much voting power these founders have in their own companies:
Elon Musk (SpaceX): 82%+
Mark Zuckerberg (Meta): ~61%
Elon Musk (Tesla): ~20%
Jensen Huang (NVIDIA): ~3.6%
Elon and Zuck made sure they retained significant control over the companies they built.
Jensen spent over 30 years building NVIDIA into one of the most important companies on the planet, yet never secured that kind of voting protection.
Of course, that doesn’t excuse his mistakes.
But it’s sad to see how easily a founder can lose control of the company he spent his entire life building.
Thoughts from my Shanghai field trip
Sentiment among Greater China investors was far worse than I expected. Unlike the US market, A shares had not seen a single rebound since July. Still holding Chinese and Hong Kong AI stocks that had crashed and Korean memory stocks that had collapsed, investors kept pressing me on when memory would bounce back.
Many investors said the AI trade was over, and a few said they doubted whether even 2027 capex, not just 2028, would go ahead as planned.
Overall, the bearish sentiment was so strong that even I, a bull before coming to China, started to waver.
It got to the point where, faced with the bear case that memory prices could start falling in the third quarter of next year, I took it seriously and couldn't refute it.
As the conversations went on, I gradually began to understand where their pessimism was coming from. In the end, they were all asking the same question.
So what comes after AI coding? What will be the next engine of ARR for Anthropic and OpenAI?
Several candidates came up. AI4S, AI chip design, and so on... But none of them clearly convinced investors.
I couldn't offer a clear answer either. Sure, there are rumors that Anthropic has made major breakthroughs in both AI4S and AI chip design, but what we actually want to know is when any of that turns into money.
With AI4S, there could be beneficiaries right away, such as companies that sell lab monkeys or diagnostic kits. In fact, there already are. But while we can trade those names in the short term, what we ultimately want to know is whether Anthropic can make money by developing drugs. That is still several years away.
As for AI chip design, the TAM itself is hard to calculate, and even if you do calculate it, it is hard to monetize. That makes it difficult for the market to believe in.
Honestly, this feels like the hardest market in history. If a company isn't making money, selling makes sense. But right now the market sells even when companies are making money, even when they beat, and even when they say demand will stay solid all the way through 2028 and 2029.
So my recommendation to investors was to buy memory, and within memory, DRAM. Not because I have conviction in the DRAM cycle, but because I see it as a relatively defensive choice within the AI trade.
Memory stocks are now trading close to 1x P/B on 2027 BPS. Historically, memory stocks have staged technical rebounds when they fell below 1x P/B.
And even if the broader market falls further from here, the three DRAM makers will keep buying back their own shares. I thought that would guarantee a certain level of buying and concluded the stocks would have downside support.
Of course, I don't know whether memory stocks can break above their previous highs. That depends on whether sentiment across the broader market recovers. But I think there is a fairly good chance that memory outperforms the rest of the AI trade.
So value guys and shorts got pasted last October, macro guys were annihilated in March, AI maxis blew up with Leo in July, rates got owned in September, everybody lost during October, and now entire pod shops are liquidating. At what point does the carnage spill over to indices?
let's be honest. if OpenAI and Anthropic went bankrupt tomorrow. What would happen
GPUs would get much cheaper. you could run Kimi and DeepSeek and get 90% of your work done just fine
you might even be able to afford rent in SF
THIS. IS. NOT. A. DRILL. 🚨
I’m spending < $40k a year on tokens/subscriptions from Anthropic, Loveable, Grok Bot, 11labs and a long tail of other products—and getting $2m+ in value @launch and @twistartups
We’ve automated at least seven full-time jobs at this point while becoming 100% faster at the average task we do
If you haven’t used the last 90 days' worth of products for at least 30 hours a week you have no idea how profound this software is
The last 90 days, things have changed more than in the last two years
Many of the things i spent hours on last year now take minutes… and the results are mind blowing
So let me get this straight...
The market is selling off AI stocks bc the largest and smartest companies in the world decided to go negative fcf for a few quarters
Because they believe investing in AI infra is better than holding cash or buying back their stock
Yet the market is selling all the stocks that receive the capital from the largest and smartest companies in the world to build out their AI infra
Ok cool, got it
Excuse me while I go pile my capital into more AI stocks
I used to think $200/mo plans were heavily subsidized. Now I think most people are burning their money buying these plans because they don't know a simple fact:
Turns out a good harness (like Claude Code) + Uncensored Qwen 3.8 Flash Next easily covers 80% of your needs. You can run it on local hardware and liberate your agents. And for just a few dollars a month, you can use Opus to write the plan for Qwen to implement.
$200/mo experience at only $20/mo Claude plan.
Oct 20: "I've never seen anything grow this fast."
Nov 1: "The cost of serving Grok Bot was unsustainably high relative to revenue."
Turns out the growth hack was selling a dollar for fifty cents.