@antpalkin
Account based inEurope!
About this account
- Account based in
- Europe
- Connected via
- Europe App Store
X says this location may be affected by a proxy or VPN.
Account-level information from X, not a live location or the device used for a specific post.
27 | research AI since 2021 | now looking for edge on quant trading with AI
dm open
Joined September 2021
- Tweets14.3K
- Following297
- Followers19.3K
- Likes42.8K
cvxv666 retweeted
jane street just signed a $13B bill for computers. five years. not traders, not offices. computers
if you have never heard the name: they are the most profitable trading firm on earth.
$39.6 billion last year out of 3,500 people. that is $11.3 million per head, seven times what a goldman banker brings in
and that firm looked at its own people and decided the machines were the better bet. $19B on ai compute this year alone
what do the machines do all day? they test trading ideas. thousands of them, all night, killing the bad ones before a single dollar moves
that is the whole job now, and it is the one part retail skips. 89% of retail traders lost money last year skipping it
same loop, your own trade ideas, somebody else paying the gpu bill. it is all in the post below
the chips came from a company worth $1.38 billion last october and about $30 billion today. the smart money is not hedging on this one
Jane Street just signed a $13,000,000,000 GPU contract to run trading strategies nobody has written yet. Five years. About $19 billion committed to AI compute this year alone.
The firm made $39.6 billion last year with 3,500 traders. Now it rents an AI factory to out-test them.
So what does $13 billion buy a trading firm?
The loop. Trade idea, code, 10-year backtest, live on the broker, autopsy, again, faster than a room full of quants can do it by hand.
3 weeks before that contract, Princeton, Stanford and Ant Group published the proof that the loop runs with nobody in the chair.
Their agent rewrites its own trading hypotheses after every backtest. Sharpe of 2.50 on US stocks after costs. Green every year from 2021 to 2025.
The paper is free. The GPUs are not.
Unless somebody already paid for yours. Entrée Capital and a group of hedge fund managers funded Horizon, and the compute behind your backtests sits on their bill.
So bring the dumbest trade idea you own. The one you never coded because it was not worth a weekend.
It comes back with 5 years of history behind it, a score out of 100, and the exact rule that was losing the money: join.horizon.trade/cvxv666
The chips came from a company worth $1.38 billion last October and about $30 billion today.
89% of retail traders lost money last year placing trades by hand. Jane Street signed for 5 years. Your first test takes 11 minutes.
cvxv666 retweeted
Jane Street just signed a $13,000,000,000 GPU contract to run trading strategies nobody has written yet. Five years. About $19 billion committed to AI compute this year alone.
The firm made $39.6 billion last year with 3,500 traders. Now it rents an AI factory to out-test them.
So what does $13 billion buy a trading firm?
The loop. Trade idea, code, 10-year backtest, live on the broker, autopsy, again, faster than a room full of quants can do it by hand.
3 weeks before that contract, Princeton, Stanford and Ant Group published the proof that the loop runs with nobody in the chair.
Their agent rewrites its own trading hypotheses after every backtest. Sharpe of 2.50 on US stocks after costs. Green every year from 2021 to 2025.
The paper is free. The GPUs are not.
Unless somebody already paid for yours. Entrée Capital and a group of hedge fund managers funded Horizon, and the compute behind your backtests sits on their bill.
So bring the dumbest trade idea you own. The one you never coded because it was not worth a weekend.
It comes back with 5 years of history behind it, a score out of 100, and the exact rule that was losing the money: join.horizon.trade/cvxv666
The chips came from a company worth $1.38 billion last October and about $30 billion today.
89% of retail traders lost money last year placing trades by hand. Jane Street signed for 5 years. Your first test takes 11 minutes.
jane street just signed a $13B bill for computers. five years. not traders, not offices. computers
if you have never heard the name: they are the most profitable trading firm on earth.
$39.6 billion last year out of 3,500 people. that is $11.3 million per head, seven times what a goldman banker brings in
and that firm looked at its own people and decided the machines were the better bet. $19B on ai compute this year alone
what do the machines do all day? they test trading ideas. thousands of them, all night, killing the bad ones before a single dollar moves
that is the whole job now, and it is the one part retail skips. 89% of retail traders lost money last year skipping it
same loop, your own trade ideas, somebody else paying the gpu bill. it is all in the post below
the chips came from a company worth $1.38 billion last october and about $30 billion today. the smart money is not hedging on this one
Jane Street just signed a $13,000,000,000 GPU contract to run trading strategies nobody has written yet. Five years. About $19 billion committed to AI compute this year alone.
The firm made $39.6 billion last year with 3,500 traders. Now it rents an AI factory to out-test them.
So what does $13 billion buy a trading firm?
The loop. Trade idea, code, 10-year backtest, live on the broker, autopsy, again, faster than a room full of quants can do it by hand.
3 weeks before that contract, Princeton, Stanford and Ant Group published the proof that the loop runs with nobody in the chair.
Their agent rewrites its own trading hypotheses after every backtest. Sharpe of 2.50 on US stocks after costs. Green every year from 2021 to 2025.
The paper is free. The GPUs are not.
Unless somebody already paid for yours. Entrée Capital and a group of hedge fund managers funded Horizon, and the compute behind your backtests sits on their bill.
So bring the dumbest trade idea you own. The one you never coded because it was not worth a weekend.
It comes back with 5 years of history behind it, a score out of 100, and the exact rule that was losing the money: join.horizon.trade/cvxv666
The chips came from a company worth $1.38 billion last October and about $30 billion today.
89% of retail traders lost money last year placing trades by hand. Jane Street signed for 5 years. Your first test takes 11 minutes.
Jane Street just signed a $13,000,000,000 GPU contract to run trading strategies nobody has written yet. Five years. About $19 billion committed to AI compute this year alone.
The firm made $39.6 billion last year with 3,500 traders. Now it rents an AI factory to out-test them.
So what does $13 billion buy a trading firm?
The loop. Trade idea, code, 10-year backtest, live on the broker, autopsy, again, faster than a room full of quants can do it by hand.
3 weeks before that contract, Princeton, Stanford and Ant Group published the proof that the loop runs with nobody in the chair.
Their agent rewrites its own trading hypotheses after every backtest. Sharpe of 2.50 on US stocks after costs. Green every year from 2021 to 2025.
The paper is free. The GPUs are not.
Unless somebody already paid for yours. Entrée Capital and a group of hedge fund managers funded Horizon, and the compute behind your backtests sits on their bill.
So bring the dumbest trade idea you own. The one you never coded because it was not worth a weekend.
It comes back with 5 years of history behind it, a score out of 100, and the exact rule that was losing the money: join.horizon.trade/cvxv666
The chips came from a company worth $1.38 billion last October and about $30 billion today.
89% of retail traders lost money last year placing trades by hand. Jane Street signed for 5 years. Your first test takes 11 minutes.
cvxv666 retweeted
A Wall Street research head told CNBC that within a year most trades will not be placed by a person.
They will be placed by the AI agents people are already wiring into their brokerage accounts.
His words: everybody gets their own family office, working 24/7, awake or asleep.
Retail places about 2 trades a month today. With agents they expect about 20 a day.
Robinhood already opened the door. Since May 27 you can plug Claude, ChatGPT or Grok straight into a trading account that has its own money and a kill switch.
Public and Podium Markets are building their own. Coinbase and Kraken are in the same race.
One guy in the report is 29. He told Claude to act as a hedge fund analyst and go find undervalued stocks.
92% of institutional volume is already automated. Retail sits at 45%.
89% of retail traders lost money last year doing it by hand.
So the agent is not the hard part anymore. Giving it your money before anything tested it is.
That part is Horizon. Type the trade idea in plain English, it writes the strategy, backtests 5 years in about 12 seconds, then goes live on your broker in about 90.
Free trial, no code: join.horizon.trade/cvxv666
Full CNBC report is below. 345,000 people have watched it. 8 minutes.
A Wall Street research head told CNBC that within a year most trades will not be placed by a person.
They will be placed by the AI agents people are already wiring into their brokerage accounts.
His words: everybody gets their own family office, working 24/7, awake or asleep.
Retail places about 2 trades a month today. With agents they expect about 20 a day.
Robinhood already opened the door. Since May 27 you can plug Claude, ChatGPT or Grok straight into a trading account that has its own money and a kill switch.
Public and Podium Markets are building their own. Coinbase and Kraken are in the same race.
One guy in the report is 29. He told Claude to act as a hedge fund analyst and go find undervalued stocks.
92% of institutional volume is already automated. Retail sits at 45%.
89% of retail traders lost money last year doing it by hand.
So the agent is not the hard part anymore. Giving it your money before anything tested it is.
That part is Horizon. Type the trade idea in plain English, it writes the strategy, backtests 5 years in about 12 seconds, then goes live on your broker in about 90.
Free trial, no code: join.horizon.trade/cvxv666
Full CNBC report is below. 345,000 people have watched it. 8 minutes.
cvxv666 retweeted
every finance bro watching a hedge fund owner fire his entire team on CNBC and pay 4 chatbots $40,000 a year to do the same $5,000,000 job
CNBC just aired a report on a 55-year-old trader who fired his entire hedge fund and replaced it with 4 AI agents.
Payroll went from $5,000,000 a year to $40,000.
He ran that fund for 8 years with 7 or 8 people in New York, Hong Kong and California.
April 2025: he sent his investors a goodbye letter and gave the money back.
10 months ago he opened Claude, pasted a price chart into it and asked what it saw.
That was the whole beginning.
Those 4 agents now cover markets 24/7 and make him 10x more productive than the humans did. His number.
One junior quant costs $400,000 to $650,000 a year. His whole firm runs on a tenth of that.
89% of retail traders lost money last year doing this by hand.
You do not have to build any of it. Horizon already did, one AI trading engine backed by Entrée Capital and a group of hedge fund managers.
Type the trade idea in plain English. 5 year backtest in about 12 seconds, live on your broker in about 90.
Free trial, no code: join.horizon.trade/cvxv666
He needed 10 months and a dead fund to get his first agent. The median here is 11 minutes.
CNBC's report is below. He explains the whole thing in 4 minutes.
cvxv666 retweeted
a guy with zero engineers replaced an entire hedge fund with 4 AI agents, and the whole thing started in a chat window
CNBC put him on air for it last week.
$5,000,000 a year in payroll, 7 or 8 people across new york, hong kong and california, all of it now running on $40,000.
he says he is 10x faster than they were
jane street pays 3,500 people $2.68m a head to run that same loop
89% of retail traders lost money last year doing it by hand
and he is not some 22 year old dev. he opened claude 10 months ago, pasted in a price chart and asked what it saw
then he kept exactly one job for himself in the whole firm
the whole story and the exact pipeline he built is in the post below
you are trading against all of it either way
CNBC just aired a report on a 55-year-old trader who fired his entire hedge fund and replaced it with 4 AI agents.
Payroll went from $5,000,000 a year to $40,000.
He ran that fund for 8 years with 7 or 8 people in New York, Hong Kong and California.
April 2025: he sent his investors a goodbye letter and gave the money back.
10 months ago he opened Claude, pasted a price chart into it and asked what it saw.
That was the whole beginning.
Those 4 agents now cover markets 24/7 and make him 10x more productive than the humans did. His number.
One junior quant costs $400,000 to $650,000 a year. His whole firm runs on a tenth of that.
89% of retail traders lost money last year doing this by hand.
You do not have to build any of it. Horizon already did, one AI trading engine backed by Entrée Capital and a group of hedge fund managers.
Type the trade idea in plain English. 5 year backtest in about 12 seconds, live on your broker in about 90.
Free trial, no code: join.horizon.trade/cvxv666
He needed 10 months and a dead fund to get his first agent. The median here is 11 minutes.
CNBC's report is below. He explains the whole thing in 4 minutes.
cvxv666 retweeted
CNBC just filmed a hedge fund where every employee is an AI agent.
Payroll: $40,000 a year, all 4 of them. His last team cost $5,000,000 and burned him out of the business.
Watch him introduce the staff.
> Houston runs the place.
> Doocey is the red team. His only job is to break every trade idea before money touches it.
> Steffi, yes, Steffi Graf, marks up the charts. >Desmond runs the quant strategies over the weekend.
The human kept one job. He calls it the meat in the chair. Pressing the button.
7 or 8 people in New York, Hong Kong and California could not cover a crypto market that trades at 3am.
4 bots do. He started them on Claude Opus 4.6 and they have not slept since. 10x the output, his number, not mine.
527,000 people watched this in 8 days. Your timeline skipped it.
His forecast for Wall Street, on camera: one hedge fund manager, 1 or 2 humans under him, a swarm of agents under them. And for himself: "Maybe someday old BK will just have to be at the beach"
A hedge fund manager wrote his investors a goodbye letter. The crypto market that never closes had broken him.
17 months later CNBC filmed his new fund. 4 AI agents, zero humans on payroll, $40,000 a year all in. Nobody in it sleeps.
The letter is still on X. 769 views.
You run the same f*cking shift he quit. Hyperliquid at 3am, a Bitcoin alert in bed. You just never had $5,000,000 of payroll to prove it wasn't working.
From the letter: "Markets that never close have taken a toll that I could only fully recognize once I paused."
From the CNBC clip: "If it's 3 in the morning and I have an idea, I can just ping them. I don't have to worry about waking somebody up in the middle of the night, which I may or may not have done in the past."
7 or 8 people in New York, Hong Kong and California. $5,000,000 a year. Could not cover 3am.
4 bots, started on Claude Opus 4.6. $40,000 a year. Never off. 10x the output, his number.
> Houston is mission control. Every bot reports to him.
> Steffi, yes, Steffi Graf, marks up the charts.
>Desmond runs the quant strategies over the weekend.
> Doocey is the red team. Every thesis gets attacked before a dollar moves. Two Sigma keeps 1,700 people on that job.
Behind them sits what he calls the corporate brain. Every trade, every email, every piece of research is a node, wired to everything else in the firm "almost like neurons".
"Take a staff of 100, you've got a staff of 1,000"
"I've talked to people working at big hedge funds now. They're doing something like this."
527,000 people watched his clip in 8 days. Brian Kelly kept one job in it. He calls it the meat in the chair.
The market still never closes. He just isn't the one awake for it anymore.
cvxv666 retweeted
CNBC just aired a report on a 55-year-old trader who fired his entire hedge fund and replaced it with 4 AI agents.
Payroll went from $5,000,000 a year to $40,000.
He ran that fund for 8 years with 7 or 8 people in New York, Hong Kong and California.
April 2025: he sent his investors a goodbye letter and gave the money back.
10 months ago he opened Claude, pasted a price chart into it and asked what it saw.
That was the whole beginning.
Those 4 agents now cover markets 24/7 and make him 10x more productive than the humans did. His number.
One junior quant costs $400,000 to $650,000 a year. His whole firm runs on a tenth of that.
89% of retail traders lost money last year doing this by hand.
You do not have to build any of it. Horizon already did, one AI trading engine backed by Entrée Capital and a group of hedge fund managers.
Type the trade idea in plain English. 5 year backtest in about 12 seconds, live on your broker in about 90.
Free trial, no code: join.horizon.trade/cvxv666
He needed 10 months and a dead fund to get his first agent. The median here is 11 minutes.
CNBC's report is below. He explains the whole thing in 4 minutes.
CNBC just filmed a hedge fund where every employee is an AI agent.
Payroll: $40,000 a year, all 4 of them. His last team cost $5,000,000 and burned him out of the business.
Watch him introduce the staff.
> Houston runs the place.
> Doocey is the red team. His only job is to break every trade idea before money touches it.
> Steffi, yes, Steffi Graf, marks up the charts. >Desmond runs the quant strategies over the weekend.
The human kept one job. He calls it the meat in the chair. Pressing the button.
7 or 8 people in New York, Hong Kong and California could not cover a crypto market that trades at 3am.
4 bots do. He started them on Claude Opus 4.6 and they have not slept since. 10x the output, his number, not mine.
527,000 people watched this in 8 days. Your timeline skipped it.
His forecast for Wall Street, on camera: one hedge fund manager, 1 or 2 humans under him, a swarm of agents under them. And for himself: "Maybe someday old BK will just have to be at the beach"
every finance bro watching a hedge fund owner fire his entire team on CNBC and pay 4 chatbots $40,000 a year to do the same $5,000,000 job
CNBC just aired a report on a 55-year-old trader who fired his entire hedge fund and replaced it with 4 AI agents.
Payroll went from $5,000,000 a year to $40,000.
He ran that fund for 8 years with 7 or 8 people in New York, Hong Kong and California.
April 2025: he sent his investors a goodbye letter and gave the money back.
10 months ago he opened Claude, pasted a price chart into it and asked what it saw.
That was the whole beginning.
Those 4 agents now cover markets 24/7 and make him 10x more productive than the humans did. His number.
One junior quant costs $400,000 to $650,000 a year. His whole firm runs on a tenth of that.
89% of retail traders lost money last year doing this by hand.
You do not have to build any of it. Horizon already did, one AI trading engine backed by Entrée Capital and a group of hedge fund managers.
Type the trade idea in plain English. 5 year backtest in about 12 seconds, live on your broker in about 90.
Free trial, no code: join.horizon.trade/cvxv666
He needed 10 months and a dead fund to get his first agent. The median here is 11 minutes.
CNBC's report is below. He explains the whole thing in 4 minutes.
a guy with zero engineers replaced an entire hedge fund with 4 AI agents, and the whole thing started in a chat window
CNBC put him on air for it last week.
$5,000,000 a year in payroll, 7 or 8 people across new york, hong kong and california, all of it now running on $40,000.
he says he is 10x faster than they were
jane street pays 3,500 people $2.68m a head to run that same loop
89% of retail traders lost money last year doing it by hand
and he is not some 22 year old dev. he opened claude 10 months ago, pasted in a price chart and asked what it saw
then he kept exactly one job for himself in the whole firm
the whole story and the exact pipeline he built is in the post below
you are trading against all of it either way
CNBC just aired a report on a 55-year-old trader who fired his entire hedge fund and replaced it with 4 AI agents.
Payroll went from $5,000,000 a year to $40,000.
He ran that fund for 8 years with 7 or 8 people in New York, Hong Kong and California.
April 2025: he sent his investors a goodbye letter and gave the money back.
10 months ago he opened Claude, pasted a price chart into it and asked what it saw.
That was the whole beginning.
Those 4 agents now cover markets 24/7 and make him 10x more productive than the humans did. His number.
One junior quant costs $400,000 to $650,000 a year. His whole firm runs on a tenth of that.
89% of retail traders lost money last year doing this by hand.
You do not have to build any of it. Horizon already did, one AI trading engine backed by Entrée Capital and a group of hedge fund managers.
Type the trade idea in plain English. 5 year backtest in about 12 seconds, live on your broker in about 90.
Free trial, no code: join.horizon.trade/cvxv666
He needed 10 months and a dead fund to get his first agent. The median here is 11 minutes.
CNBC's report is below. He explains the whole thing in 4 minutes.
CNBC just aired a report on a 55-year-old trader who fired his entire hedge fund and replaced it with 4 AI agents.
Payroll went from $5,000,000 a year to $40,000.
He ran that fund for 8 years with 7 or 8 people in New York, Hong Kong and California.
April 2025: he sent his investors a goodbye letter and gave the money back.
10 months ago he opened Claude, pasted a price chart into it and asked what it saw.
That was the whole beginning.
Those 4 agents now cover markets 24/7 and make him 10x more productive than the humans did. His number.
One junior quant costs $400,000 to $650,000 a year. His whole firm runs on a tenth of that.
89% of retail traders lost money last year doing this by hand.
You do not have to build any of it. Horizon already did, one AI trading engine backed by Entrée Capital and a group of hedge fund managers.
Type the trade idea in plain English. 5 year backtest in about 12 seconds, live on your broker in about 90.
Free trial, no code: join.horizon.trade/cvxv666
He needed 10 months and a dead fund to get his first agent. The median here is 11 minutes.
CNBC's report is below. He explains the whole thing in 4 minutes.
CNBC just filmed a hedge fund where every employee is an AI agent.
Payroll: $40,000 a year, all 4 of them. His last team cost $5,000,000 and burned him out of the business.
Watch him introduce the staff.
> Houston runs the place.
> Doocey is the red team. His only job is to break every trade idea before money touches it.
> Steffi, yes, Steffi Graf, marks up the charts. >Desmond runs the quant strategies over the weekend.
The human kept one job. He calls it the meat in the chair. Pressing the button.
7 or 8 people in New York, Hong Kong and California could not cover a crypto market that trades at 3am.
4 bots do. He started them on Claude Opus 4.6 and they have not slept since. 10x the output, his number, not mine.
527,000 people watched this in 8 days. Your timeline skipped it.
His forecast for Wall Street, on camera: one hedge fund manager, 1 or 2 humans under him, a swarm of agents under them. And for himself: "Maybe someday old BK will just have to be at the beach"
cvxv666 retweeted
A hedge fund manager wrote his investors a goodbye letter. The crypto market that never closes had broken him.
17 months later CNBC filmed his new fund. 4 AI agents, zero humans on payroll, $40,000 a year all in. Nobody in it sleeps.
The letter is still on X. 769 views.
You run the same f*cking shift he quit. Hyperliquid at 3am, a Bitcoin alert in bed. You just never had $5,000,000 of payroll to prove it wasn't working.
From the letter: "Markets that never close have taken a toll that I could only fully recognize once I paused."
From the CNBC clip: "If it's 3 in the morning and I have an idea, I can just ping them. I don't have to worry about waking somebody up in the middle of the night, which I may or may not have done in the past."
7 or 8 people in New York, Hong Kong and California. $5,000,000 a year. Could not cover 3am.
4 bots, started on Claude Opus 4.6. $40,000 a year. Never off. 10x the output, his number.
> Houston is mission control. Every bot reports to him.
> Steffi, yes, Steffi Graf, marks up the charts.
>Desmond runs the quant strategies over the weekend.
> Doocey is the red team. Every thesis gets attacked before a dollar moves. Two Sigma keeps 1,700 people on that job.
Behind them sits what he calls the corporate brain. Every trade, every email, every piece of research is a node, wired to everything else in the firm "almost like neurons".
"Take a staff of 100, you've got a staff of 1,000"
"I've talked to people working at big hedge funds now. They're doing something like this."
527,000 people watched his clip in 8 days. Brian Kelly kept one job in it. He calls it the meat in the chair.
The market still never closes. He just isn't the one awake for it anymore.
CNBC just filmed a hedge fund where every employee is an AI agent.
Payroll: $40,000 a year, all 4 of them. His last team cost $5,000,000 and burned him out of the business.
Watch him introduce the staff.
> Houston runs the place.
> Doocey is the red team. His only job is to break every trade idea before money touches it.
> Steffi, yes, Steffi Graf, marks up the charts. >Desmond runs the quant strategies over the weekend.
The human kept one job. He calls it the meat in the chair. Pressing the button.
7 or 8 people in New York, Hong Kong and California could not cover a crypto market that trades at 3am.
4 bots do. He started them on Claude Opus 4.6 and they have not slept since. 10x the output, his number, not mine.
527,000 people watched this in 8 days. Your timeline skipped it.
His forecast for Wall Street, on camera: one hedge fund manager, 1 or 2 humans under him, a swarm of agents under them. And for himself: "Maybe someday old BK will just have to be at the beach"
A hedge fund manager wrote his investors a goodbye letter. The crypto market that never closes had broken him.
17 months later CNBC filmed his new fund. 4 AI agents, zero humans on payroll, $40,000 a year all in. Nobody in it sleeps.
The letter is still on X. 769 views.
You run the same f*cking shift he quit. Hyperliquid at 3am, a Bitcoin alert in bed. You just never had $5,000,000 of payroll to prove it wasn't working.
From the letter: "Markets that never close have taken a toll that I could only fully recognize once I paused."
From the CNBC clip: "If it's 3 in the morning and I have an idea, I can just ping them. I don't have to worry about waking somebody up in the middle of the night, which I may or may not have done in the past."
7 or 8 people in New York, Hong Kong and California. $5,000,000 a year. Could not cover 3am.
4 bots, started on Claude Opus 4.6. $40,000 a year. Never off. 10x the output, his number.
> Houston is mission control. Every bot reports to him.
> Steffi, yes, Steffi Graf, marks up the charts.
>Desmond runs the quant strategies over the weekend.
> Doocey is the red team. Every thesis gets attacked before a dollar moves. Two Sigma keeps 1,700 people on that job.
Behind them sits what he calls the corporate brain. Every trade, every email, every piece of research is a node, wired to everything else in the firm "almost like neurons".
"Take a staff of 100, you've got a staff of 1,000"
"I've talked to people working at big hedge funds now. They're doing something like this."
527,000 people watched his clip in 8 days. Brian Kelly kept one job in it. He calls it the meat in the chair.
The market still never closes. He just isn't the one awake for it anymore.
A hedge fund manager wrote his investors a goodbye letter. The crypto market that never closes had broken him.
17 months later CNBC filmed his new fund. 4 AI agents, zero humans on payroll, $40,000 a year all in. Nobody in it sleeps.
The letter is still on X. 769 views.
You run the same f*cking shift he quit. Hyperliquid at 3am, a Bitcoin alert in bed. You just never had $5,000,000 of payroll to prove it wasn't working.
From the letter: "Markets that never close have taken a toll that I could only fully recognize once I paused."
From the CNBC clip: "If it's 3 in the morning and I have an idea, I can just ping them. I don't have to worry about waking somebody up in the middle of the night, which I may or may not have done in the past."
7 or 8 people in New York, Hong Kong and California. $5,000,000 a year. Could not cover 3am.
4 bots, started on Claude Opus 4.6. $40,000 a year. Never off. 10x the output, his number.
> Houston is mission control. Every bot reports to him.
> Steffi, yes, Steffi Graf, marks up the charts.
>Desmond runs the quant strategies over the weekend.
> Doocey is the red team. Every thesis gets attacked before a dollar moves. Two Sigma keeps 1,700 people on that job.
Behind them sits what he calls the corporate brain. Every trade, every email, every piece of research is a node, wired to everything else in the firm "almost like neurons".
"Take a staff of 100, you've got a staff of 1,000"
"I've talked to people working at big hedge funds now. They're doing something like this."
527,000 people watched his clip in 8 days. Brian Kelly kept one job in it. He calls it the meat in the chair.
The market still never closes. He just isn't the one awake for it anymore.
cvxv666 retweeted
A CNBC Fast Money trader shut down his $5,000,000 a year hedge fund team and rebuilt the whole f*cking firm on 4 AI agents. New bill: $40,000 a year.
He gave CNBC the tour 10 days ago, bot by bot.
You are still telling yourself AI trading needs a desk.
His 4 agents run the loop every quant desk runs:
idea -> code -> backtest -> live -> autopsy.
Horizon runs that loop from one text box.
Kelly closed a hedge fund to get his agents. Yours come on a free trial, no code, first backtest 11 minutes after signing up for the median trader.
Bring the one trade idea you never coded: join.horizon.trade/cvxv666
Type the sentence. Rules written, 5 years backtested in about 12 seconds, live on your broker 90 seconds later. Backed by Entrée Capital and hedge fund managers.
> $5,000,000 a year for 7 or 8 humans and a New York office
> $40,000 a year $40,000 a year for Houston, Steffi, Desmond and Doocey, compute included
> $400,000 to $650,000 a year for one junior quant at a big fund
I read it twice, sure I had missed a zero. That is not a hedge fund budget. That is a used car.
Steffi marks up the charts, Desmond runs the quant strategies, Doocey is the red team that attacks every thesis, Houston is mission control. Kelly keeps the final call, 10x more productive by his count.
It will not flatter you. I fed it the MACD crossover every $2,000 course sells. 227 trades, Sharpe -0.48, -0.80%. Red on the first pass, not 2 weeks into my real money. Then it rewrote the losing rule.
Kelly kept the final call. The $5,000,000 was everything before it.
A CNBC Fast Money trader shut down his $5,000,000 a year hedge fund team and rebuilt the whole f*cking firm on 4 AI agents. New bill: $40,000 a year.
He gave CNBC the tour 10 days ago, bot by bot.
You are still telling yourself AI trading needs a desk.
His 4 agents run the loop every quant desk runs:
idea -> code -> backtest -> live -> autopsy.
Horizon runs that loop from one text box.
Kelly closed a hedge fund to get his agents. Yours come on a free trial, no code, first backtest 11 minutes after signing up for the median trader.
Bring the one trade idea you never coded: join.horizon.trade/cvxv666
Type the sentence. Rules written, 5 years backtested in about 12 seconds, live on your broker 90 seconds later. Backed by Entrée Capital and hedge fund managers.
> $5,000,000 a year for 7 or 8 humans and a New York office
> $40,000 a year $40,000 a year for Houston, Steffi, Desmond and Doocey, compute included
> $400,000 to $650,000 a year for one junior quant at a big fund
I read it twice, sure I had missed a zero. That is not a hedge fund budget. That is a used car.
Steffi marks up the charts, Desmond runs the quant strategies, Doocey is the red team that attacks every thesis, Houston is mission control. Kelly keeps the final call, 10x more productive by his count.
It will not flatter you. I fed it the MACD crossover every $2,000 course sells. 227 trades, Sharpe -0.48, -0.80%. Red on the first pass, not 2 weeks into my real money. Then it rewrote the losing rule.
Kelly kept the final call. The $5,000,000 was everything before it.
cvxv666 retweeted
the GPT-6 Astra brain, my neural copy of 169 winning Robinhood traders, got hands this week. 4 new commands in TG bot, one smarter alert.
/follow. every fill of unipcs or DumbCrayonEater in your chat 20 seconds after the block. 15x faster than a screener. 3 wallets free, 50 on PRO
/paper. $100 into every alert, out an hour later. 88% went above the call, 34% held it to the hour, +$2,807 since sep 11. worst trade -$99, best +$2,494
/invite. your own link. 7 days of PRO for both of you. 4 invites, free month
/webhook. alerts posted straight into your own bot the second they fire. 566K requests a day, 48 bots read over 1,000 times each.
at $20 a month in $FOMOBRAIN that's $960 burned every month. one more turn of the flywheel.
the alert got smarter too. a deployer with two seeded tokens doesn't get a third push. three wallets buying in the same minute count as one opinion. somebody spent $12,386 on 144 fake buys to make it fire. 12 waves, 0 got through.
god bless
I used GPT-6 Astra to build a neural copy of 169 winning memecoin traders, then handed it money.
+$2,088 in a week. ×2.2 on capital. Nobody touched it.
82% of its calls went above entry. It posts every one of them to X and Telegram by itself.
Anyone can get the same calls.
139,454 fills. 6,289 tokens. 536 wallets scored. The top one is up $10.8M in 30 days. Still $0 a month in API bills.
Every 20 seconds:
> READS every fill on Robinhood Chain, 20 seconds behind the block
> CHECKS the receipt on each one. 1 in 11 "smart money buys" on this chain is fake, it zeroes them
> SCORES every trader 0 to 100 the way the 169 would, reasoning written out
> FIRES when trusted wallets pile into one coin inside 30 minutes. 66 calls last week, 54 above entry
> TRADES its own bursts. $100 a ticket, 17 tickets, +$2,088 in 5 days
> POSTS every call to Telegram 20 seconds after the fill and to X an hour later
CATGPT ran ×27 after the call.
musebook did ×18 on a $100 ticket and covered all 9 losers of the week.
Jane Street has a desk of humans reading trades like this and the results never leave the building. This one puts them on a public timeline an hour later.
The site, the bot, the X account and the code are all in my profile.
Next the algo goes live on a public wallet, and the flywheel gets its second turn.
the GPT-6 Astra brain, my neural copy of 169 winning Robinhood traders, got hands this week. 4 new commands in TG bot, one smarter alert.
/follow. every fill of unipcs or DumbCrayonEater in your chat 20 seconds after the block. 15x faster than a screener. 3 wallets free, 50 on PRO
/paper. $100 into every alert, out an hour later. 88% went above the call, 34% held it to the hour, +$2,807 since sep 11. worst trade -$99, best +$2,494
/invite. your own link. 7 days of PRO for both of you. 4 invites, free month
/webhook. alerts posted straight into your own bot the second they fire. 566K requests a day, 48 bots read over 1,000 times each.
at $20 a month in $FOMOBRAIN that's $960 burned every month. one more turn of the flywheel.
the alert got smarter too. a deployer with two seeded tokens doesn't get a third push. three wallets buying in the same minute count as one opinion. somebody spent $12,386 on 144 fake buys to make it fire. 12 waves, 0 got through.
god bless
I used GPT-6 Astra to build a neural copy of 169 winning memecoin traders, then handed it money.
+$2,088 in a week. ×2.2 on capital. Nobody touched it.
82% of its calls went above entry. It posts every one of them to X and Telegram by itself.
Anyone can get the same calls.
139,454 fills. 6,289 tokens. 536 wallets scored. The top one is up $10.8M in 30 days. Still $0 a month in API bills.
Every 20 seconds:
> READS every fill on Robinhood Chain, 20 seconds behind the block
> CHECKS the receipt on each one. 1 in 11 "smart money buys" on this chain is fake, it zeroes them
> SCORES every trader 0 to 100 the way the 169 would, reasoning written out
> FIRES when trusted wallets pile into one coin inside 30 minutes. 66 calls last week, 54 above entry
> TRADES its own bursts. $100 a ticket, 17 tickets, +$2,088 in 5 days
> POSTS every call to Telegram 20 seconds after the fill and to X an hour later
CATGPT ran ×27 after the call.
musebook did ×18 on a $100 ticket and covered all 9 losers of the week.
Jane Street has a desk of humans reading trades like this and the results never leave the building. This one puts them on a public timeline an hour later.
The site, the bot, the X account and the code are all in my profile.
Next the algo goes live on a public wallet, and the flywheel gets its second turn.