cvxv666
@antpalkin • 19,160 subscribers
27 | research AI since 2021 | now looking for edge on quant trading with AI
Shorts
Videos
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"
cvxv6661,678,582 次观看 • 1 个月前
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.
cvxv666995,254 次观看 • 1 个月前
Citadel fired their best quant. He rebuilt their entire algo with Claude Fable 5 in 48 hours - and he's up $430,000 trading it against them. He didn't take a single file. He didn't need to - ten years of that logic lived in his head, and you can't raid a memory. Wallet proof: Here's the engine MiroFish runs - and it's rigged in his favor. Picture a Galton board: a ball dropping through eight rows of pegs, bouncing left or right at random. One ball is chaos. Thousands of balls always fall into the same bell curve. That's the law he weaponized. Every ball is one trade. Each row is a volatility gate - news, liquidations, order-book flow, things nobody controls. On a fair board every gate is a 50/50 coin flip. His model tilts each one to 0.54 - four cents of edge that only shows up when fair value splits from the book. Four cents sounds like nothing. Compound it through eight gates, thirty-two thousand times, and the whole bell shifts right of breakeven: 71% of trades land green. $93 of edge per trade. $430k across the distance. Watch the win rate converge in real time - it swings between 50 and 85% for the first few dozen trades, then locks on 0.71 and never leaves. He doesn't predict a single trade. One trade is a coin flip. Eighteen thousand is mathematics. They thought firing him protected the edge. They just handed it a grudge. Copy the wallet quietly out-trading a $60B fund before they connect the dots:
cvxv6662,465,639 次观看 • 4 个月前
Grok Bot by Elon Musk just replaced a $294,000 Wall Street research desk. Six AI agents, $200 a month, doing the work BlackRock pays people six figures for. The guy who built it published every prompt. What that desk cost before the bots: > $27,000 a year for the Bloomberg terminal > $22,000 for Refinitiv > $50,000 for sell-side research > $15,000 for AlphaSense > $180,000 for the junior analyst who reads all of it until 2am so somebody senior can trade on it before the open $294,000 a year. The agents cost $2,400. Same brief, 122 times cheaper, and it is there before the open. Six agents split that desk. Each one gets its own computer in the cloud, and they all write into the same vault. > FILINGS reads every 10-K, 10-Q and 8-K on a 100-ticker watchlist overnight and flags the ugly parts: going concern language, auditor changes, restatements > EARNINGS reads the call transcript within 24 hours and tells you if the CFO got quieter than he was last quarter > SECTOR does one pass per sector and picks up the rule change nobody read > INSIDER catches Form 4 buys over $1 million and new positions from Two Sigma, Third Point and D.E. Shaw the day the 13Fs land > CHATTER counts mentions on X and flags any ticker three standard deviations louder than its own 30-day normal > CHIEF OF STAFF reads the other five at 5:30am, bins anything only one of them flagged, and emails you the ranked brief at 6 One analyst covers 30 names. Six agents cover 3,000 and finish before the open. You show them the job once. They repeat it every night with your laptop shut. Every research floor on Wall Street was priced on reading being slow and people being expensive. Both of those stopped being true this month. So you wake up, read for five minutes, and know what moved on every name you hold. That used to be somebody's entire job, and he got a bonus for it. The article below is the step by step guide to building the whole AI trading machine. Save & read it, you will want it open while you build.
cvxv666826,122 次观看 • 2 个月前
I used GPT-6 Astra to make a neural copy of every winning memecoin trader on Robinhood and wire them into one brain. 374 wallets. 116,420 fills. 5,918 tokens. Every buy, sell and cost basis in one database. The top score printed $19.5M in 30 days. It's live and it's free: Fomo app leaderboard hands you a PnL number. Some of those are one lucky ticket, and 1 in 11 "buys" credited to a top wallet on this chain is somebody else's swap with a famous name on it. Radar eats all of that for breakfast: > RESOLVER finds the real wallet behind every fomo profile. 101 for 101 against fomo's verified list > TAPE reads the chain every 20 seconds, fill in the database before you've refreshed the chart > PROVENANCE checks the receipt on every fill. Fake buy, dust, seeded token, all three get zeroed > JUDGE is GPT-6 Astra. Reads the whole book with cost bases, scores every trader 0 to 100, reasoning shown > BURST fires when four 80+ wallets ape the same token inside 30 minutes. 5 a day, 39% go 2x > EXITS shows the same wallets leaving. Nobody else publishes this FLYBRAIN, this week. Eight trusted wallets in by 20:18. Alert fired 20:18. Pool didn't open until 20:51. It ran 27x. CRUMBS, day one of the exit feed. 22 trusted buyers at breakfast, 19 gone by dinner, every exit on the feed. Jump Crypto keeps a floor of people to read flow like this. The whole Radar roster is two RPC calls + AI. No insider feeds it, no paid group, and I can't push a coin into it any more than you can. Seed a token into the tracked wallets and it vanishes from every feed. Just the model, tens of thousands of lines of code, and a database built one block at a time off the chain. $0 a month. Paying for data made it worse. Next: a real balance behind it, and the copy starts trading like the 169 wallets it was built from. We got here first. God bless us.
cvxv666443,106 次观看 • 1 个月前
GPT6-Astra turned $500 into $5,110 on my trading account last night. Six AI agents, 0.4 seconds a round, and I was asleep for all of it. 255 setups came up. 1,151 configs died. One trade reached my money. The agent holding the veto killed 4 of the last 5. I don't get a vote either. You have the same model open in a tab right now. You're using it to ask whether the market looks bullish. Six seats, one evening, and the whole roster fits in six prompts. The names and the rules they run on are in the post below ↓
cvxv666270,351 次观看 • 1 个月前
I gave Elon's Grok Bot $1,000 and one rule. Grow it or I delete you. 15 hours later it is holding $3,900. Every trade of it landed on a public address. No private account and no screenshots, nothing here sits behind my word. Every cycle it: → reads the news and the filings before the rest of the market gets to them → checks who is putting money behind the story, not who is talking about it → scores the idea and sizes it against a hard risk cap → writes the exit before it takes the entry → logs the fill, the slippage and the fee separately → reads its own losing trades at night and rewrites the rules by morning It does not care what you trade. Gold, oil, a stock you inherited or a coin that launched an hour ago, the rules are the same. Setup took one evening. Create the chief of staff, hand the rest job titles the way you brief new hires, run one task on your screen while they watch, connect Telegram. Nothing to install past the app and nobody to wait on. Morgan Stanley pays a person six figures to do one slice of that job. Mine costs $300 a month, and it does not sleep, does not get bored, and has no opinion about last week. 15 hours in and it sits at $3,900. What do you think it has in a week? The whole build is going up soon. Every job title, every prompt, so anyone can stand one up.
cvxv666342,243 次观看 • 1 个月前
Chinese student received a $4M investment the very next day after building his MiroFish simulation system. Now you can simulate anything. Build a virtual world, tweak the parameters, and watch how things plays out in every possible scenario. Catch every weakness and surprise twist before it even happens. Infinite runs. Infinite data. SPX reaction to news? Bro already simulated it and printed $200k, see quoted post. Trump does something dumb -> load the data, run the sim, see exactly how voters lose their minds. You can literally simulate the entire universe with this thing. The process is simple: simulate -> analyze -> improve -> repeat. Full guide + pipeline in quoted post, if you want to run a similar simulation on your own data.
cvxv6661,147,552 次观看 • 7 个月前
Elon's Grok Bot runs an entire trading desk for $300 a month. The person who does that job at Goldman starts at $200,000. Mine took $1,000 to $3,900 on day one, out of a wallet anyone can open and read. Most of the build is not about making it trade. It is about stopping it. Every agent's job description ends with a list of what it may never do without asking me first. One of the six is not allowed to trade at all. Its entire job is killing what the other five found, and its no beats every score on the desk. The one that makes the final call never touches the market either. You could have this running by Friday. Most people reading this will not, and in a few months they will be paying somebody who did. The whole build is in the post below ↓ Save it while it still sounds strange. Give it a quarter and this is just how a retail account works.
cvxv666208,448 次观看 • 1 个月前
$FOMOBRAIN is live. the neural copy of 169 winning memecoin traders on Robinhood has a ticker now, and the ticker is how the brain pays for itself. CA: 0x5cdf61bef103d9b9fffe2b6edb6aab541ceec1ac what's already done: - 116,420 fills, 5,918 tokens, 374 traders, 36 days of tape, all of it read straight off the chain - the brain itself. a neural copy of the 169 wallets that keep winning, built on GPT-6 Astra - live. bursts, fresh launches, exits, a score on every trader with the reasoning shown - 24/7 watch. the chain gets read every 20 seconds, every fill gets checked, nobody touches it - the whole code on github. 17,476 lines, MIT what's next, in order: - TG bot goes public. signals straight from the brain, in your pocket - a public trading algo on a real balance. public wallet, every trade readable on chain, the brain picks and i don't - the flywheel the flywheel, plain: signals in the bot are sold for $FOMOBRAIN. every token spent on them is burned every trade of $FOMOBRAIN pays creator fees. those fees are the algo's trading deposit, and the algo trades it on the brain's own signals 50% of what the algo makes goes back into the deposit. the other 50% buys $FOMOBRAIN off the market and burns it the finished algo gets sold to a closed group for $FOMOBRAIN, or rented for it. burned either way so every road ends at the same place. more subscribers, more burn. more volume, bigger deposit, more profit, bigger buyback, more burn no numbers promised. i don't know how fast this spins yet, i know which way it spins god bless
cvxv666159,084 次观看 • 1 个月前
Jane Street spent $6 billion and 4,032 GPUs on one thing: AI agents that loop through trading strategies 100x faster than any human. Generate, backtest, kill, repeat - thousands a week, until one survives data it's never seen. That loop is why they made $39.6 billion last year while 89% of retail went broke trading one idea they never stress-tested. The edge was never a smarter idea. It was volume. They test 100 strategies for every 1 you test by hand, and a second agent kills the 97 that only look good on old data before a dollar goes live. Now the entire loop fits in a chat box. Horizon takes one plain-English sentence and spins up the same two agents. One builds your strategy, one tries to kill it - backtests 5 years in seconds, and only what survives goes live on your exchange. The $150 billion setup. Free while beta open. Bookmark this & read the full breakdown below - the exact two-agent loop, step by step, and who just handed it to retail.
cvxv666318,185 次观看 • 3 个月前
Trump's crypto advisor aka "Czar" David Sacks is insider trading on Polymarket. He knows in advance when Trump will say something related to crypto. $187k profit. 90% win rate in the mentions category. He makes $30-50k from each of Trump's public appearances. His profile: Peak boomer hobby - fun and profitable. Longest win streak: 51 in a row. Bookmark his wallet: if this guy bets that Trump will say something, then Trump will definitely say it. Pro tip: skip his NBA trades - dude yolo's on his favorit team (they are losers)
cvxv666681,326 次观看 • 8 个月前
Ken Griffin, CEO of Citadel, said: "I'm fairly depressed watching AI do a week of PhD work in a few hours" He's worth $51.2 billion and 20% of every stock trade in America goes through his firm. The money isn't what's bothering him. Citadel pays $400,000 to $650,000 a year for that work. New trading AI agent does it now, on a free trial, for people who have never written a line of code. It went live this summer and most of your timeline still hasn't noticed. You type one sentence in plain English. AI agents loop takes it from there: > an AI agent turns your sentence into a real trading strategy > backtests it across 5 years of data in about 12 seconds > scores it 0 to 100 and shows you the exact spot where it leaks > kills the versions that only look good on paper > rebuilds what's left and runs the loop again > stress-tests the survivor on years it has never seen, then puts it live on your exchange in about 90 seconds That list is the entire job description of a junior quant. At a desk, one properly tested idea burns about $87,500 in salary time, and most of what they build is dead by week six. Two clicks and this AI agent is testing your own idea. Costs you nothing and it runs the whole thing by itself: Griffin gets to be depressed about it. He already owns the desk. You just got handed one. Bookmark & read full story of how they got an AI to build trading strategies and kill its own bad ones is in the article below.
cvxv666225,653 次观看 • 2 个月前
GPT6-Astra ran six AI agents on my trading account last night. One pot that only grows, one gate, 1,151 of their own ideas killed before a dollar of mine moved. $500 in. $5,110 in the pot by morning. Nobody lost their seat. > PROBE finds volume that moves before price > PRISM checks who paid for the spike > SCALE does the money math and sizes what's left > LIMIT marks where the trade stops being a trade > ARCHIVE keeps the ledger and remembers what the rest forgot > EINSTEIN is the gate. Red light and the round ends. PROBE brought 25 candidates. PRISM threw out 20. EINSTEIN vetoed 4 of the 5 left. One setup reached my money. Then it went straight down. The ladder was already there. Six buy orders written before the first candle, $220 at the top, $986 at the bottom. All six filled on the way down. I'd have been closing the tab at level three. D. E. Shaw pays a room to argue a trade down to size. Mine does it in 0.4 seconds and never asks for a bonus. None of the six knows more than the chatbot on your phone. None gets to be right on its own. Setup is one evening. Name the six, one job each, twice as long on the gate. The whole desk exists to throw work away. 1,151 dead ideas last night, one trade. How it's wired is in the article below.
cvxv666125,551 次观看 • 1 个月前
"Long oil the second Trump says something bad about Iran" Someone typed that into an AI agent. It pulled the data, wrote the code, ran a 10-year backtest, and shipped it live to their broker. Minutes. Every trader has an idea like that rotting in their head - never tested, because coding it wasn't worth a weekend. One that got typed in: buy stocks that gap up 2%, hold 30 days. Backtest came back +562%, profit factor 3.27. One click, live on Alpaca. The $650k quant job, now a text box. First 10,000 traders. Go type the idea you've been sitting on for two years. Demo's below.
cvxv666245,608 次观看 • 3 个月前
A one-man trading hedge fund now costs about $100 a month in AI subscriptions. Five PhDs on a Citadel desk cost $270,000 a month, and that is the cheap end. That buys one tested idea a month. Man Group's AI writes hundreds of new signals a week, on a floor where the humans used to manage twenty in a quarter. Fifty AI agents do the reading and the writing. One of them exists only to destroy what the others build, and that one is the reason any of it works. Every job on that payroll is now an agent you can run yourself: > AI research agents reading filings, transcripts, options flow and on-chain data overnight, fifty personas at once, none of them allowed to see what the crowd is pricing > coding agent that turns one sentence of plain English into entry rules, exit rules, sizing and risk limits, then debugs itself until it runs > backtest agent that replays five years of that idea in about 12 seconds > breaker agent, an AI whose only job is killing the thing, at double the trading costs and in the ten worst markets that asset has ever seen > critic agent that reads your journal and names the mistake you keep making Nobody on that list asks for a bonus in January. $200 million is already run this way on one platform where 52,000 people mostly sit and watch the machine work. The full map of what is built and what is missing is in the article below. Bookmark & read it If you're not yet using AI in your trading and investing.
cvxv666173,490 次观看 • 2 个月前
Jane Street made $39.6 billion last year with just 3,500 people. That's $11.3M per employee - 7x Goldman. They didn't do it with more humans. They spent $6 billion on AI and 4,032 GPUs in a Texas data center to make each quant 10x faster. There's now a tool that lets anyone test same tradings strategys in 90 seconds - no coding, no finance degree, just plain English -> Here's what's happening inside the top firms: Man Group ($150B) built an AI on Claude that writes and tests strategies on its own - hundreds a week. A human team tests 20 in a quarter. Bridgewater runs a $2B AI fund making "alpha uncorrelated to what our humans do" The edge was never the idea. It was speed. They test 100 strategies for every 1 you test by hand. Horizon hands that exact speed to regular people. Plain English in. Tested system out. Save this. Re-read it when the waitlist is closed.
cvxv666337,654 次观看 • 5 个月前
A 29-year-old sales consultant from Oklahoma quit his job thanks to AI - now he makes his director’s entire annual salary in just two weeks. $306,000 in profit over the past month. He replaced an entire team of quants with Claude and a swarm of agents - and built his own ETH price simulation engine. His wallet: Each of the 6 agents validates its own trading decision, and thanks to their precision, he’s pulling in at least $15,000 a day. The agent system coverage speed is far higher than any top-tier trading team. They collect data 24/7 and run simulations using that data in the MiroFish engine - completely autonomously. They remember every pattern, every market reaction, every trading signal. This is a new milestone in algo trading. Every trade is a perfect cycle. Every dollar earned is pure exploitation of market inefficiency. He doesn’t predict the future - the math already knows it. He just reads the numbers right and takes the money. Save this post if you actually want to learn how to build something like this. Or just skip homework and start copying every one of his trades right now - that’s the easiest and most profitable route I’m on:
cvxv666337,817 次观看 • 5 个月前
A 35-year-old accountant from New Orleans left his job and spent a full month in isolation with Claude. The result? He made $45,000 in a single day. 300 hours of meticulous work - and the perfect BTC trading algorithm was ready. If he keeps cooking like this, he’s hitting over $1,000,000 in a single month. His wallet: He selected username nsh91qaz - an ironic nod to his 1991 birth year, an age when most people believe it’s too late to change their lives. But he changed anyway. I ran a backtest of his strategy using Claude + Nautilus via PyPI. Results genuinely shocked me - mechanics are understandable to pretty much anyone. The real alpha is in the numbers under the hood. That’s what lets you pull $45k per day with pure math. I simulated every single one of his trades and broke down every transaction: 75 markets, 72 fills, 85.1% win rate, Sharpe ratio 4.21. All run on the Nautilus-core broker simulator with 41.8 GB of parquet data in DuckDB. Every trade is a perfect cycle. Every dollar earned is pure exploitation of market inefficiency. He doesn’t predict the future - the math already knows it. He just reads the numbers right and takes the money Brier-loss ensemble: 400 trees · lr 0.03, walk-forward validation with Sharpe 4.21 ± 0.08. Save this post if you actually want to learn how to build something like this. Or just skip the homework and start copying his trades right now - that’s the easiest and most profitable route I’m on:
cvxv666341,820 次观看 • 5 个月前
GPT6-Astra is running Squid Game on my trading account. Six agents, one pot, $500 in and $6,300 out by morning while I slept. The rules I gave them are the show's rules. The pot only grows. Whoever is wrong the most gets switched off and their seat goes to somebody new. Jump Trading pays people to sit on a desk and say no. Mine does it for free and doesn't go home. Six players: > 456 hunts contracts where volume moves before the price does > 067 reads the crowd and flags which spike somebody paid for > 218 does the money math and sizes whatever is left standing > 240 marks the level where the trade stops being a trade > 001 has sat through every round and is the only one who talks to me > THE FRONT MAN calls red light and that ends the conversation Red light on the first setup of the night. I wrote that rule myself and I was still sour about it at midnight. Then the chart went vertical without us in it. That's the exact minute I'd have chased it and donated the difference. The desk didn't move. It sat under red light until the price came back down to 240's level, and took the entry there instead. By morning the pot read $6,300 and nobody had been eliminated. None of them is smarter than the model you already pay for. They're small enough to only be wrong about one thing at a time, and there's always somebody sitting between them and the button. Setup is an evening. Number them, give each one job, write the Front Man's rules twice as carefully as the rest, then go to bed. The expensive part of a trading desk was never the ideas. It was paying six people to disagree with each other at 3am. Full build is in the article below.
cvxv666100,114 次观看 • 1 个月前