
I like to mimic the brain with technology. | trader | founder @playfair_app | https://xitter.cf/t.co/hZy4wnLrIf |
The World
Joined January 2021
- Tweets8.8K
- Following1K
- Followers2K
- Likes22.4K
Pinned Tweet
Fantastic reply Parag. Never thought I'd say it but this is great form
Parag Agrawal (@paraga) on how he dealt with pressure during @elonmusk’s takeover of Twitter:
“Through the Twitter acquisition, I used to track my sleep score. And I was sleeping fine.”
“If you know what you must do, you do it. You don't have to feel pressure.”
“The world can come at you, but if you know what you're capable of & you have a slightly long-term view.. you don't have to feel pressure. You have to perform every day, which is a completely different thing.”
Absolutely awful take by an awful investor who should not be listened to despite having AI in his name.
Happy SK Hynix earnings day to everyone. Wouldn't be right if we didn't drop our pre-call thoughts. View article
Terrible News
Oh no. Bad news for the memory bulls. SK Hynix only made $62 billion dollars of profit in one quarter. For perspective, that’s about 7 years worth of Tesla profits. Or 15 years worth of PLTRs current yearly profits. Or 6 years of AMDs current yearly profits. All in one quarter. Not one year, one quarter. But it really doesn’t matter, because memory is cyclical. And cyclical money is like Monopoly money. All of these really really smart people on X taught me that cyclical money doesn’t count. The memory bears are going to teach those memory bulls a lesson. The cyclical guys are so smart. They know what’s best. $60 billion dollars of profit in one quarter doesn’t matter when it’s from selling memory stuff that is sooooooo cyclical. Oh, and BTW, SK said their HBM will be sold out until 2030. It may be longer, but they can’t see out past that. But don’t worry, memory is so cyclical that I’m sure it will peak soon. I’m so happy I follow so many smart guys that taught me about the cyclicality of memory. Good thing they were all predicting memory stocks to crash. Even the really smart guys with those charts and squiggly lines. They knew it. I can’t wait to see MU at $600 a share soon like they predict. And DRAM at $25 a share. lol. 🤡
Fuckin cyclical clowns out there. When will they figure out that this time really is different? Ai never existed during the past cycles. So, by definition, it is different. That’s ok. I hope they figure it out when MU passes $5000 a share. lol.
This may be a big deal still deciding. Will test if it actually works on non cache input
Alright I finally have more to share on my personal project!
Resin is a runtime for building fast LLM inference engines.
The idea is to make all the generic parts of inference reusable, and leave only the model-specific forward pass + kernels to be written for each architecture / hardware target.
To build an engine:
1. Import Resin's core.
2. Have an agent write a forward pass for the model you’re interested in, for your specific hardware (GPU or TPU or AMD), any kernels you want.
Then, boom, you have a standalone inference engine.
—
I've been hacking over the past month (early post: xitter.cf/ekzhang1/status/208950…), and recently this week I got a lot of features + parallelisms working, inspired by DeepInfer.
For DeepSeek-V4.1-Flash, it's nearly as fast as DeepInfer (within 5-10%), while being self-contained and supporting other models, plus metrics + HTTP server + prefix cache.
Yesterday I wrote a skill for authoring models and put Codex to work on it overnight. The agent built a GLM-5.3-Flash engine with MTP, reaching up to 60,000 input tok/s and 16,000 output tok/s on B200 (EP4).
So far inference "vibe coding" has not been very reliable or successful. I think it's because people start from scratch each time and reinvent lots of tricky things on the scheduler and server layers. But with a shared core, we can automate lots of inference work.
The tricky bit is writing a fast forward pass. But Resin turns that into a standalone engine by just importing a library, and it becomes a verifiable goal for the AI.
Code: github.com/ekzhang/resin
My thoughts exactly. The denominator can get very small from an AI perspective with reasonable chance of algorithmic success as a result.
Absolutely agree with Dario here. Skepticism is cyclical across domains as tech broadens across advancement periods. Therein is the opportunity.
Tremendous clip about Dario’s reaction when he realized he invented a God: 🔥
“I was super freaked out... I said, ‘My God, you just scale this thing up by another factor of a million and it will be better than humans at everything.’
- Upon witnessing GPT-2 responding to scaling laws (more data = better model) in 2019
We also learn Dario and Daniela hold grudges (watch to the end):
“We remember…”
Upon witnessing GPT-2’s generalizability across domains, where it wasn’t yet great at anything but it was capable at many things:
“This is the defining economic and scientific moment of the modern age. This is going to determine the fate of nations…and the future of humanity.
“It really scared me. And maybe 10% of the people around me had exactly the same reaction, and it was often the smartest ones.”
Incredibly well said. We will hopefully one day look back on Toronto's current scenario as a hazy dream.
Dear Toronto,
We need to talk.
I was born here. I learned to drive on the Gardiner, I know the exact smell of the TTC in July, and I’ve spent more of my life than I’d like to admit arguing about whether the east end is better than the west end. This city is in my bones. Which is why it hurts so much to say this: I’m done.
Olivia Chow just got re-elected. Again. And I can’t pretend this is still the place I fell in love with.
Look, I tried. I defended the bike lanes while sitting in traffic that now moves at the speed of a progressive policy paper. I paid the property tax increases with a smile and a slight twitch in my left eye.
Toronto, I’ve seen you at your best. Summer nights at the island, that one perfect slice at a place that closed forever during covid in 2020, the feeling when the streetcar actually shows up. But I’ve also watched you slowly become a place where the most common conversation is either “how much is your rent?” or “did you hear about the latest thing that happened on the subway?” And now the person in charge of steering us out of it just got another four years.
I can’t keep doing this. I can’t keep explaining to out-of-town friends why a city this rich still can’t pick up its trash, clear a sidewalk, or house its people without turning every solution into a morality play. I can’t keep pretending that freezing fares while everything else goes up is some kind of genius economic strategy. And I definitely can’t keep watching the same debates about “community safety” end with a new working group and a press release.
So I’m leaving. I’m taking my Canadian Tire money, my residual fondness for the skyline, and my deep, abiding love of complaining about the weather, and I’m going somewhere that might actually change something when people are unhappy.
You’ll always be my first love. But right now you’re the ex who keeps texting “we can make this work” while clearly dating the same person who broke your heart the first time.
Take care of yourself. Or don’t. Apparently most of you are fine either way.
Sincerely,
A native who checked the real estate listings south of the border
P.S. If you ever want to try again, you know where to find me. Just don’t bring Olivia.
(I asked Grok to write this breakup letter for me because I couldn't even be bothered enough about Toronto any more to do it myself. I did make a few edits, however.)
A Surprisingly Accurate Account of Modern Times
PE was a path to guaranteed money for long enough that people assumed it would always be a path to guaranteed money, making it increasingly prestigious.
But prestige is a indicator of where the world went, not a forecast for where the world is going.
youngmoney.co/p/prestige-tra…
This is obvious. And they will absolutely Mogg
i have a strong feeling that all the frontier labs which exists today will start directly competing with companies across different sectors. probably pharma companies, defence companies so on and so forth. as open-source models get even better, the revenue growth will die down. in next few years it might decline too. so the only way, they keep growing is by creating new IP, not with just better frontier models but also with groundbreaking discoveries.
So cool
Sequoia is running an experiment
We’re hosting a poker bot tournament, with rules for the age of AI
There will be prizes, an incredible guest list (ex @polynoamial), and some plot twists
The brainchild of my partner @abhishekm1636
Excited for this!
Brandon Marsh retweeted
I don't know how many people realize that if Elon Musk pulls off what he's building, it will be the biggest social reset in human history.
To understand why, you need to understand a concept almost nobody really gets: rent.
Yesterday I was talking about it with someone. I told them something simple: the moment a rent sets in, whatever kind it is, it gets paid for. Not necessarily in dollars. It gets paid for in stagnation.
So what is a rent?
A rent is income that doesn't come from the value you create, but from the position you hold. You don't win because you're better. You win because nobody is allowed to come and beat you.
The mechanism is almost always the same. A company succeeds. Then, instead of continuing to innovate, it goes for regulatory capture: licenses, standards, quotas, approvals, lobbying. It turns its lead into a wall. And behind that wall, new entrants can't get in.
But they're the ones who innovate. They're the ones who drive prices down, raise quality, create new wealth. When you block them, nothing dramatic happens. Nothing happens at all. Prices stop falling. Quality stalls. The pie stops growing. And when the pie stops growing, the only way to get richer is to take someone else's slice.
This is the point almost nobody sees: what people blame on "capitalism" (frozen inequality, a broken social ladder, the feeling that the game is rigged from the start) are the effects of that stagnation. The symptoms of rent. Not of capitalism. Capitalism is competition. Rent is its negation.
Now, Elon Musk.
He said that by 2036, money won't matter anymore. Most people read it as science fiction. I read it as an implicit critique of our entire monetary system.
His reasoning: AI and robots will produce more than we can possibly consume. So prices will collapse. Deflation, not inflation.
And this is where it gets interesting. For decades, the role of central banks has been to stabilize prices. In practice: to stop them from falling. Yet in an economy that innovates, the natural direction of prices is down. Look at what a computer cost thirty years ago.
Preventing prices from falling was the worst thing we could have done. Because it created an entire generation of rentiers. People who got rich not by creating something, but by owning something at the right time. An apartment. A plot of land. An asset shielded from falling prices by money printing and by rules that choke supply.
In a world not captured by the state, house prices would be extremely low.
And who would stay rich? The people who keep creating value. The ones whose wealth comes from the valuation of their company, because their company keeps innovating. Wealth you have to re-earn every single year.
That's where Elon's line really makes sense. In a world of deflationary abundance, human life would look like a video game. Or a game of chess.
If you're good and you start at the bottom, you can rise very, very fast. In a few years.
If you're bad and you inherit the inertia of a position you don't know how to maintain, against people who keep innovating, you fall very, very fast too.
No more rent. No more entitlement. Just your ability to create value, put back on the line every day.
It would be the greatest equalizer in human history.
Not equality of outcomes imposed from above. Equality of the playing field, imposed by abundance.
I think Elon Musk is reinventing the rules of capitalism. Or rather, bringing them back to their original form: the Austrian school. Real capitalism. Value creation as the only legitimate source of wealth.
And at the end of the road, something neither rent nor redistribution has ever achieved: the end of poverty on a global scale.
Still confused why big labs don't have ultra granular trace inspection
Today, @JudgmentLabs opens for general access.
We believe long running agent systems will generate more exhaust data than the world can currently process.
Hidden within agent traces are rich records of how to make these systems better: what users ask for, when agents stumble, and how agents interact with the external world.
Judgment is the most performant monitoring and improvement layer for long-horizon agents.
We’re excited to see what you build!
Was looking for something like this!!
Today, @JudgmentLabs opens for general access.
We believe long running agent systems will generate more exhaust data than the world can currently process.
Hidden within agent traces are rich records of how to make these systems better: what users ask for, when agents stumble, and how agents interact with the external world.
Judgment is the most performant monitoring and improvement layer for long-horizon agents.
We’re excited to see what you build!
Brandon Marsh retweeted
Talk to anyone inside OpenAI and they will tell you that Sam and his top engineers use Ultraftast powered by $CBRS. People have gotten in trouble internally for abusing Ultrafast at 10x the cost. Jane St has attempted to monopolize the limited supply. Feldman screwed up on the
I think the opposite
Nick is an online course salesman. Not a trader.
My predictions over the following months:
1. Stock market will get hurt bad when the music stops on data centers, models, and chips.
Too much froth, too high of investment, too many companies will fail. The switching costs are low. I don't know how these companies will become profitable. Its a mega bubble.
The rest of the economy is already in a deep recession. Real pain will set in when the public markets dive.
2. Power is the bottleneck.
AI is burning an absurd amount of power. The cost of electricity is skyrocketing which is making the cost of everyday life more expensive and hurting quality of life. I'm against data centers because of the power, not the water.
The grid is extremely fragile and the weakness will be exposed in the coming months.
3. Subsidies are skewing the market.
AI isn't cheap. The prices to use it are artificially low because of subsidies all the way up the chain.
You pay $20 or $200 a month but the true costs are 10x That spread is venture capital money being incinerated at scale.
When they raise the price, the users will disappear and investors will hold the bag.
3. A big security breach is coming.
A bank, a certain model, something is going to rock our world.
People are letting bots log in to their banks, their email and everything else. Big trouble coming.
4. Grokbot is exceptional.
Its the first model that is adding significant value to my everyday life. I don't know why anybody would use a different personal assistant.
5. AI is insanely high maintenance
Billions are being poured into training the next model. They day it comes out its already behind. China has cheaper electricity and more chips.
The chips they bought will be obsolete in 18 months.
Companies are stuck paying big money every month to maintain and update all of their AI tools. Its a mess.
6. The power companies will renege on their deals.
Political pressure will unwind all the unsustainable power deals these data centers and LLMs have negotiated.
Power costs will explode. Then comes the choice:
Do we jack up costs for our AI / compute and watch everybody switch to the next model or do we incinerate more VC capital? Nobody wins.
7. Credit will be the first domino to fall.
Data centers are already struggling to finance the GPUs over 10 years when they will be obsolete in 18 months. We'll end up with a shit load of half built data centers who can't negotiate power rates or buy chips because nobody will let them borrow money.
The money will get cold feet as soon as the markets tank and the show will be over.
8. It'll be a dark, dark period in economic markets.
The people who own assets that make real money (real estate or businesses) will win.
The people who plan ahead and avoid massive leverage will win.
The people with cash on their balance sheet will accumulate more assets and generate exceptional wealth.
--
Don't listen to me. I'm just a bear. Not investment advice. Good luck!



















