Carver's banner

Carver

@carverfomo • 17,182 subscribers

tracking probabilities | trading narratives

Shorts

Every night at 2 AM someone clicks a button and makes $49,000. The wallet is 57 days old. ClawdBot flagged it before my morning coffee. Now I can't stop checking. I showed anoin123 to a friend who trades professionally. He asked if I photoshopped the screenshot. $1,613,408 profit on Polymarket. Not in a year. In under two months. That's hedge fund performance from a browser tab. Dug into the timeline. The entries don't fire randomly. Almost every trade lands between 2-4 AM EST when volume is dead and most traders are asleep. Whatever signal this wallet follows, it only speaks at night. Changed the ClawdBot filter: profit above $500K, wallet age under 90 days. Three results. Total. On the entire platform. Under 20 wallets in Polymarket history have ever crossed $1M. This one did it faster than most people finish a trading course. My quant friend stopped laughing. Now he's asking how to set up alerts. → No - US strikes Iran by February 13, 2026

Every night at 2 AM someone clicks a button and makes $49,000. The wallet is 57 days old. ClawdBot flagged it before my morning coffee. Now I can't stop checking. I showed anoin123 to a friend who trades professionally. He asked if I photoshopped the screenshot. $1,613,408 profit on Polymarket. Not in a year. In under two months. That's hedge fund performance from a browser tab. Dug into the timeline. The entries don't fire randomly. Almost every trade lands between 2-4 AM EST when volume is dead and most traders are asleep. Whatever signal this wallet follows, it only speaks at night. Changed the ClawdBot filter: profit above $500K, wallet age under 90 days. Three results. Total. On the entire platform. Under 20 wallets in Polymarket history have ever crossed $1M. This one did it faster than most people finish a trading course. My quant friend stopped laughing. Now he's asking how to set up alerts. → No - US strikes Iran by February 13, 2026

44,402 次观看

$28,669 in a single day on Bitcoin Up or Down. Today’s round: price to beat 93,506 BTC trading well below the line while odds barely adjust. That gap is exactly what blubberbuster keeps exploiting on Polymarket. He trades only Bitcoin, but the model compounds hard his monthly total is now 304,490.38. Same pattern every session: → waits for the micro drift → enters before the market recalculates → exits as the line finally catches up When the feed moves faster than the pricing, this is what happens.

$28,669 in a single day on Bitcoin Up or Down. Today’s round: price to beat 93,506 BTC trading well below the line while odds barely adjust. That gap is exactly what blubberbuster keeps exploiting on Polymarket. He trades only Bitcoin, but the model compounds hard his monthly total is now 304,490.38. Same pattern every session: → waits for the micro drift → enters before the market recalculates → exits as the line finally catches up When the feed moves faster than the pricing, this is what happens.

20,441 次观看

What happens when you stop guessing and let machine learning read the market for you? $2.2M in 4 months. ilovecircle built something different on Polymarket. Not a speed bot. Not a spread farmer. An AI system that actually thinks. 1,347 predictions. 74% win rate. Biggest single hit: $258.4K. Current positions: basically zero he extracted everything. The setup: 10 machine learning models running in parallel, each trained on news feeds and social media data. They don't predict events they predict when the crowd is wrong about probabilities. Market prices an outcome at 50 cents. His ensemble says the real odds are 60%. That gap is the trade. Every week the models retrain themselves on fresh data. The edge evolves because the system never stops learning. Most traders react to headlines. This wallet front runs the market's understanding of what headlines actually mean. 51K people watching now. Most still think AI trading is a scam until they see a curve like this. → Following wallets that run AI-powered probability models is simpler with PMX.

What happens when you stop guessing and let machine learning read the market for you? $2.2M in 4 months. ilovecircle built something different on Polymarket. Not a speed bot. Not a spread farmer. An AI system that actually thinks. 1,347 predictions. 74% win rate. Biggest single hit: $258.4K. Current positions: basically zero he extracted everything. The setup: 10 machine learning models running in parallel, each trained on news feeds and social media data. They don't predict events they predict when the crowd is wrong about probabilities. Market prices an outcome at 50 cents. His ensemble says the real odds are 60%. That gap is the trade. Every week the models retrain themselves on fresh data. The edge evolves because the system never stops learning. Most traders react to headlines. This wallet front runs the market's understanding of what headlines actually mean. 51K people watching now. Most still think AI trading is a scam until they see a curve like this. → Following wallets that run AI-powered probability models is simpler with PMX.

13,072 次观看

How do you pull six figures without guessing Bitcoin’s next move? You wait for the site to be late. This wallet lives in the 15 minute BTC windows and only buys when the odds still show yesterday’s price. Bitcoin moves fast Polymarket updates slower for 30-90 seconds. That gap is the entire edge. He repeats the same entry thousands of times, taking a fixed profit each window. 2277 trades since August. Sometimes it’s a normal win sometimes one window prints $69 K. He runs the same idea on ETH and SOL, holding multiple windows at once so there’s always a lag to harvest. → No - Bitcoin above $90 K on January 21 Would you rather guess direction or bet on the site being late? PMX lets you watch and copy these entries in real time.

How do you pull six figures without guessing Bitcoin’s next move? You wait for the site to be late. This wallet lives in the 15 minute BTC windows and only buys when the odds still show yesterday’s price. Bitcoin moves fast Polymarket updates slower for 30-90 seconds. That gap is the entire edge. He repeats the same entry thousands of times, taking a fixed profit each window. 2277 trades since August. Sometimes it’s a normal win sometimes one window prints $69 K. He runs the same idea on ETH and SOL, holding multiple windows at once so there’s always a lag to harvest. → No - Bitcoin above $90 K on January 21 Would you rather guess direction or bet on the site being late? PMX lets you watch and copy these entries in real time.

12,261 次观看

$3.6M on Polymarket from one boring edge: the crowd overpays. The bait is simple: most people bet like fans. SeriouslySirius trades like a book. He sits almost only in NFL/NBA spreads because liquidity is deep enough to enter and exit without getting trapped. When the public slams the favorite, the underdog gets marked too cheap for a few minutes. He buys that cheap side, anchors his risk to the Vegas line and sells once the price snaps back. That’s why the curve looks impossible. Not because he’s always right on the final score. Because he keeps getting paid on the correction before settlement. → No - Hull vs Watford (EFL Championship, Jan 4)

$3.6M on Polymarket from one boring edge: the crowd overpays. The bait is simple: most people bet like fans. SeriouslySirius trades like a book. He sits almost only in NFL/NBA spreads because liquidity is deep enough to enter and exit without getting trapped. When the public slams the favorite, the underdog gets marked too cheap for a few minutes. He buys that cheap side, anchors his risk to the Vegas line and sells once the price snaps back. That’s why the curve looks impossible. Not because he’s always right on the final score. Because he keeps getting paid on the correction before settlement. → No - Hull vs Watford (EFL Championship, Jan 4)

11,607 次观看

Videos

Anya Rossi

sweetdream.ai

SweetDream.ai•Sponsored•Livecam

Watch Anya Live

Anya is streaming live right now! Join her private show and enjoy exclusive content.

Exclusive private shows

1.2k viewers online

Private Show

Join now for exclusive access

Free preview available • Premium content

Chinese PhD student forgot to close his GitHub before a lab meeting. Screen shared for 11 seconds. His advisor saw a repo called polymarket-sports-arbitrage. The advisor didn't report him. He asked for the wallet address. 432614799197. $2.67M profit. 2,911 predictions. All sports NFL, Premier League, Ligue 1. Account created in January 2026. $824K bet on PSG won't win → $2.28M payout. $1.13M bet on Bills vs Jaguars → $2.45M payout. $781K bet on Packers vs Bears → $1.79M payout. Million dollar positions. From a guy who lives in a campus apartment and bikes to class. The bot tracks Asian bookmaker lines that move 2-3 hours before Polymarket updates. By the time American traders wake up, the mispricing is already gone. He catches it while they sleep. The GitHub README has 127 stars now. Last line: The market doesn't know what timezone you're in. Use that. His advisor asked to invest. He said no the edge works better with less capital in the pool. 223K people watching. $214K still loaded in active positions. The lab meeting ended awkwardly. The wallet didn't care. → No 28c - Girona vs Barcelona

Chinese PhD student forgot to close his GitHub before a lab meeting. Screen shared for 11 seconds. His advisor saw a repo called polymarket-sports-arbitrage. The advisor didn't report him. He asked for the wallet address. 432614799197. $2.67M profit. 2,911 predictions. All sports NFL, Premier League, Ligue 1. Account created in January 2026. $824K bet on PSG won't win → $2.28M payout. $1.13M bet on Bills vs Jaguars → $2.45M payout. $781K bet on Packers vs Bears → $1.79M payout. Million dollar positions. From a guy who lives in a campus apartment and bikes to class. The bot tracks Asian bookmaker lines that move 2-3 hours before Polymarket updates. By the time American traders wake up, the mispricing is already gone. He catches it while they sleep. The GitHub README has 127 stars now. Last line: The market doesn't know what timezone you're in. Use that. His advisor asked to invest. He said no the edge works better with less capital in the pool. 223K people watching. $214K still loaded in active positions. The lab meeting ended awkwardly. The wallet didn't care. → No 28c - Girona vs Barcelona

1,372,814 次观看 • 5 个月前

A Chinese engineering student built a Claude powered device the size of a credit card. Raspberry Pi. Camera. One red LED. Total cost $14. His professor called it a toy. His classmates laughed. He posted a demo on YouTube and got 200 views. Then someone noticed what the device was actually connected to. A wallet. Making $11 million. In 10 months. kch123. $11,525,240 profit. 2,347 predictions. Joined June 2025. → Profile: The device does one thing. He asks it a question with his voice. Claude scans thousands of prediction markets in real time. When it finds a price that doesn't match reality the red LED lights up. Red light. He places the trade. Green result. $2M position on Seahawks. Won. $1.5M on Villarreal not winning. His professor saw the YouTube video again three months later. The demo was the same. The view count wasn't. Neither was the wallet. He asked the student what changed. The student showed him the balance. The professor didn't say anything. Sat down. Then asked if the device could run two accounts. 759K people watching. $319K still loaded in active positions. The device fits in your hand. The LED is still blinking. His classmates stopped laughing around month four. By month six two of them had built their own.

790,272 次观看 • 3 个月前

2021: My cousin had 12 GPUs mining ETH in his garage. Made $4K/month. Electricity bill: $1,800. 2025: China banned mining. GPUs worthless. He sold everything for parts. 2026: I visited him last week. Same garage. Different machines. 6 phones. 3 laptops. 1 monitor. All running the same script. Green numbers everywhere. What is this? Claude Code farm. Same idea as mining. But I don't mine coins. I farm 5 minute markets. rename. 31,221 predictions. $383K profit. Joined January 2026. Each device runs the same bot. Claude Code wrote it in one afternoon. 47 lines Python. Every 5 minutes a new BTC market opens. Bot scans. Bot enters. Bot collects. Mining needed: GPUs, cooling, electricity, luck. This needs: phone, WiFi, Claude Code. His electricity bill now: $89. His profit last month: $47K. I asked: Why multiple devices? Rate limits. Each device runs one instance. 6 devices = 6x positions per window. The garage that lost money on mining now prints money on Polymarket. I stood there for 20 minutes watching phones buzz with profits. He offered me an old iPhone. Start your own farm. 27K watching. $1,554 loaded right now. Mining died. Farming started.

2021: My cousin had 12 GPUs mining ETH in his garage. Made $4K/month. Electricity bill: $1,800. 2025: China banned mining. GPUs worthless. He sold everything for parts. 2026: I visited him last week. Same garage. Different machines. 6 phones. 3 laptops. 1 monitor. All running the same script. Green numbers everywhere. What is this? Claude Code farm. Same idea as mining. But I don't mine coins. I farm 5 minute markets. rename. 31,221 predictions. $383K profit. Joined January 2026. Each device runs the same bot. Claude Code wrote it in one afternoon. 47 lines Python. Every 5 minutes a new BTC market opens. Bot scans. Bot enters. Bot collects. Mining needed: GPUs, cooling, electricity, luck. This needs: phone, WiFi, Claude Code. His electricity bill now: $89. His profit last month: $47K. I asked: Why multiple devices? Rate limits. Each device runs one instance. 6 devices = 6x positions per window. The garage that lost money on mining now prints money on Polymarket. I stood there for 20 minutes watching phones buzz with profits. He offered me an old iPhone. Start your own farm. 27K watching. $1,554 loaded right now. Mining died. Farming started.

720,877 次观看 • 4 个月前

A Japanese TV crew filmed the CEO of a 7 billion yen fast food chain for a feature on Japan's humble corporate culture. His head office in Shinagawa had 3 desks and cost 103,000 yen a month in rent. The office was that small because a Claude agent did the procurement, pricing, and contract work that normally fills three floors. The crew showed the office. Two desks. One printer. Three employees. A rice cooker in the corner. The TV banner introduced him as the entrepreneur whose chain of snack shops sources rice directly from 312 farmers across Japan. At 0:43 he says the word fleet. He says it once. He says it without looking at the camera. The crew kept the line because they thought he meant his delivery vans. He did not mean delivery vans. He meant the fleet of Claude agents that runs every part of his company that does not require a human signature. The two desks at headquarters are for him and the CFO. The third employee is there to answer the phone. One agent reads daily yield data from 312 rice farmers and sets the next morning's wholesale price for each variety. A second agent routes 47 stores worth of inventory based on the previous day's POS data. A third agent drafts every franchise contract and every supplier renewal. He signs every morning before the office opens. Japanese commercial law is satisfied. Someone pulled the company's filings on the Tokyo commercial registry. Every contract filed in the last 14 months had been timestamped between 5:47 AM and 6:03 AM. Every supplier renewal used the same boilerplate clauses, written in slightly different prose each time. The morning shots in the TV segment showed an empty office because the agent had already done the work of 40 employees overnight. Six months ago a 14 year old in Shenzhen pushed an AI agent to GitHub. Judges said no real world application. 3,100 forks later. The CEO had been one of them. He still flies to rice paddies every spring. He still personally tastes every new variety before it gets a code in the system. He still tells investors the head office rent is 103,000 yen because it builds trust. He still has not told the farmers that the agent decided who got the new orders last quarter. The TV crew thought the rundown head office was a story about a humble entrepreneur. It was actually a story about how many employees a 7 billion yen company does not need when one CEO signs what one Claude agent writes.

A Japanese TV crew filmed the CEO of a 7 billion yen fast food chain for a feature on Japan's humble corporate culture. His head office in Shinagawa had 3 desks and cost 103,000 yen a month in rent. The office was that small because a Claude agent did the procurement, pricing, and contract work that normally fills three floors. The crew showed the office. Two desks. One printer. Three employees. A rice cooker in the corner. The TV banner introduced him as the entrepreneur whose chain of snack shops sources rice directly from 312 farmers across Japan. At 0:43 he says the word fleet. He says it once. He says it without looking at the camera. The crew kept the line because they thought he meant his delivery vans. He did not mean delivery vans. He meant the fleet of Claude agents that runs every part of his company that does not require a human signature. The two desks at headquarters are for him and the CFO. The third employee is there to answer the phone. One agent reads daily yield data from 312 rice farmers and sets the next morning's wholesale price for each variety. A second agent routes 47 stores worth of inventory based on the previous day's POS data. A third agent drafts every franchise contract and every supplier renewal. He signs every morning before the office opens. Japanese commercial law is satisfied. Someone pulled the company's filings on the Tokyo commercial registry. Every contract filed in the last 14 months had been timestamped between 5:47 AM and 6:03 AM. Every supplier renewal used the same boilerplate clauses, written in slightly different prose each time. The morning shots in the TV segment showed an empty office because the agent had already done the work of 40 employees overnight. Six months ago a 14 year old in Shenzhen pushed an AI agent to GitHub. Judges said no real world application. 3,100 forks later. The CEO had been one of them. He still flies to rice paddies every spring. He still personally tastes every new variety before it gets a code in the system. He still tells investors the head office rent is 103,000 yen because it builds trust. He still has not told the farmers that the agent decided who got the new orders last quarter. The TV crew thought the rundown head office was a story about a humble entrepreneur. It was actually a story about how many employees a 7 billion yen company does not need when one CEO signs what one Claude agent writes.

142,952 次观看 • 1 个月前

His parents paid $280K for his MIT education. Computer Science. 4 years. They wanted him to get into Google. He wrote a bot in 3 weeks. Named it pm-timezone-arbitrage. The bot made $4.4M in 3 months. Google pays $180K/year. His bot makes that in 2 weeks. 432614799197. 4,390 predictions. All sports. NFL, Premier League, Ligue 1. The strategy his parents paid $280K to teach him: track Asian bookmaker lines that move 2-3 hours before American platforms update. Bet while the US sleeps. Collect when they wake up. His mom called last week: Did you hear back from Google? He said: I'm not going to Google. She cried. Then he sent her a screenshot. → Profile: 278K watching. $1.2M loaded in active positions. He still lives in a campus apartment. Still bikes to class. The diploma cost $280K. The bot paid it back in 6 weeks. His dad asked if he should've just given him the tuition money instead. $280K education. One repo. PMX found it first.

370,253 次观看 • 4 个月前

Two Chinese developers recorded a 4 minute video showing how to build a team of 7 AI agents that replaces an entire customer support department for $50 a month. They opened a terminal. In 4 minutes they had the whole thing running. One agent classifies tickets. One reads the knowledge base. One handles billing. One watches the customer's tone. One decides when to escalate. One writes the report. One agent runs the other six. The math hit hard. 1,200 tickets a day per person. More than a call center of 8 operators. The company used to pay $25,000 to $40,000 a month in salaries. Now they pay $50 for the API. The video hit 4 million views in 72 hours. Every CEO in China was forwarding it. Every support team in America was panicking. While the West is still debating whether AI will replace workers, China just published the manual. These guys were teaching people how to fire a department. They just showed too much. Bro pause at 0:39. Ignore the guy in the blue shirt pointing. Look at the second monitor on the right. That window is not a support dashboard. That is a live wallet. gabagool22. $868,862 profit. 28,620 predictions. Joined October 2025. → They were filming a tutorial about replacing humans with cheap AI. Their own setup behind them was running an AI that replaces traders. 28,620 positions. All BTC. All 15 minute windows. All green. The comment section turned into a detective board. Someone slowed the video to 0.5x. Screenshotted every frame where the right monitor was visible. Stitched them together. Reconstructed the full wallet page from 6 seconds of background footage. Entry prices between 2 and 10 cents. Payouts in the thousands. Not one red row in 28,620 entries. Biggest single win: $4,696. From a 15 minute window on a Tuesday. The 7 support agents save a company $40K a month. The wallet on their second monitor makes that in one good week. The tutorial taught the world how to replace a customer support team. The setup behind them was already replacing the trading desk. They deleted the wallet zoom from the next upload. Too late. Someone had already screen recorded it. The clip hit Discord. Then Telegram. Then every dev forum. The tutorial got 4 million views. The zoom on their second monitor got another 800,000. 727K people watching the wallet now. The 7 agents are still answering tickets somewhere for $50 a month. The wallet behind them does not need 7 agents. It needs one. And that one was already running while they filmed.

Two Chinese developers recorded a 4 minute video showing how to build a team of 7 AI agents that replaces an entire customer support department for $50 a month. They opened a terminal. In 4 minutes they had the whole thing running. One agent classifies tickets. One reads the knowledge base. One handles billing. One watches the customer's tone. One decides when to escalate. One writes the report. One agent runs the other six. The math hit hard. 1,200 tickets a day per person. More than a call center of 8 operators. The company used to pay $25,000 to $40,000 a month in salaries. Now they pay $50 for the API. The video hit 4 million views in 72 hours. Every CEO in China was forwarding it. Every support team in America was panicking. While the West is still debating whether AI will replace workers, China just published the manual. These guys were teaching people how to fire a department. They just showed too much. Bro pause at 0:39. Ignore the guy in the blue shirt pointing. Look at the second monitor on the right. That window is not a support dashboard. That is a live wallet. gabagool22. $868,862 profit. 28,620 predictions. Joined October 2025. → They were filming a tutorial about replacing humans with cheap AI. Their own setup behind them was running an AI that replaces traders. 28,620 positions. All BTC. All 15 minute windows. All green. The comment section turned into a detective board. Someone slowed the video to 0.5x. Screenshotted every frame where the right monitor was visible. Stitched them together. Reconstructed the full wallet page from 6 seconds of background footage. Entry prices between 2 and 10 cents. Payouts in the thousands. Not one red row in 28,620 entries. Biggest single win: $4,696. From a 15 minute window on a Tuesday. The 7 support agents save a company $40K a month. The wallet on their second monitor makes that in one good week. The tutorial taught the world how to replace a customer support team. The setup behind them was already replacing the trading desk. They deleted the wallet zoom from the next upload. Too late. Someone had already screen recorded it. The clip hit Discord. Then Telegram. Then every dev forum. The tutorial got 4 million views. The zoom on their second monitor got another 800,000. 727K people watching the wallet now. The 7 agents are still answering tickets somewhere for $50 a month. The wallet behind them does not need 7 agents. It needs one. And that one was already running while they filmed.

234,605 次观看 • 2 个月前

Chinese student in my MIT dorm asked ChatGPT and Gemini the same question: How can I make money using the timezone difference between Asia and America? ChatGPT explained forex trading hours and international freelancing. Gemini suggested remote work for Asian companies while living in the US. He closed both tabs. Standard answers he'd heard a hundred times before. Then he opened Claude and asked differently. Not how to make money but who is already making money this way and can you find them. Claude took 30 seconds. Then explained that prediction markets have a gap Asian bookmaker lines move 2-3 hours before American platforms update. Someone in Beijing sees price movements that someone in New York won't see until they wake up. He asked Claude to find wallets that only trade during Beijing morning hours. Wallets that enter positions when America sleeps. Claude found one that matched perfectly. Every single trade executed between 6-9 AM Beijing time. 4,548 of them. 432614799197. $4.5M profit. Joined January 2026. Only sports NFL, Premier League, Ligue 1. The Chinese student understood immediately. While his American classmates sleep from 11 PM to 7 AM Eastern, someone in Asia has an 8 hour window where they see tomorrow's lines today. He showed me the wallet last week. I asked why he shared it. He said the edge isn't the wallet the edge is knowing what question to ask. ChatGPT and Gemini answered the question he asked. Claude answered the question he meant. 305K watching. $79K loaded in active positions.

Chinese student in my MIT dorm asked ChatGPT and Gemini the same question: How can I make money using the timezone difference between Asia and America? ChatGPT explained forex trading hours and international freelancing. Gemini suggested remote work for Asian companies while living in the US. He closed both tabs. Standard answers he'd heard a hundred times before. Then he opened Claude and asked differently. Not how to make money but who is already making money this way and can you find them. Claude took 30 seconds. Then explained that prediction markets have a gap Asian bookmaker lines move 2-3 hours before American platforms update. Someone in Beijing sees price movements that someone in New York won't see until they wake up. He asked Claude to find wallets that only trade during Beijing morning hours. Wallets that enter positions when America sleeps. Claude found one that matched perfectly. Every single trade executed between 6-9 AM Beijing time. 4,548 of them. 432614799197. $4.5M profit. Joined January 2026. Only sports NFL, Premier League, Ligue 1. The Chinese student understood immediately. While his American classmates sleep from 11 PM to 7 AM Eastern, someone in Asia has an 8 hour window where they see tomorrow's lines today. He showed me the wallet last week. I asked why he shared it. He said the edge isn't the wallet the edge is knowing what question to ask. ChatGPT and Gemini answered the question he asked. Claude answered the question he meant. 305K watching. $79K loaded in active positions.

71,859 次观看 • 4 个月前

A Chinese mathematician posted a 3 minute video on Bilibili explaining how he lost his $10,000 a month gig to AI. The model he had been training started writing harder math problems than he could invent. He admitted his own mistake in business positioning. He had spent four years hand writing PhD level math problems for Scale AI's reinforcement learning pipeline. $50 to $100 per problem. 200 problems a month. Then synthetic data killed his entire contract category. He was no longer able to invent a problem the machine could not solve. At 2:13 he says the word agent. He says it once. He never says it again in the video. The way he says it is the only thing on screen that did not come off the teleprompter. He has been recording videos off a teleprompter for three months. The teleprompter runs on the same agent that killed his Scale AI work. Every script is generated by Claude. Every word he reads to camera is the agent's. The new job is reading. Someone pulled the script repository from a Cursor instance the dev had left public. The folder was labeled bilibili-laments. Inside were 47 video scripts. All in his voice. All written by Claude. Six months ago a 14 year old in Shenzhen pushed an AI agent to GitHub. Judges said no real world application. 3,100 forks later. The mathematician had been one of them. He had wired the agent into his content pipeline the week Scale AI cut him off. He had been a PhD candidate at one of the top five Chinese math schools. He taught there for two years before going full time on Scale AI contracts. He still has the credentials. He still has the office. He just no longer writes anything. He wanted to show people how AI took his career. He accidentally showed them how AI also took his post mortem.

A Chinese mathematician posted a 3 minute video on Bilibili explaining how he lost his $10,000 a month gig to AI. The model he had been training started writing harder math problems than he could invent. He admitted his own mistake in business positioning. He had spent four years hand writing PhD level math problems for Scale AI's reinforcement learning pipeline. $50 to $100 per problem. 200 problems a month. Then synthetic data killed his entire contract category. He was no longer able to invent a problem the machine could not solve. At 2:13 he says the word agent. He says it once. He never says it again in the video. The way he says it is the only thing on screen that did not come off the teleprompter. He has been recording videos off a teleprompter for three months. The teleprompter runs on the same agent that killed his Scale AI work. Every script is generated by Claude. Every word he reads to camera is the agent's. The new job is reading. Someone pulled the script repository from a Cursor instance the dev had left public. The folder was labeled bilibili-laments. Inside were 47 video scripts. All in his voice. All written by Claude. Six months ago a 14 year old in Shenzhen pushed an AI agent to GitHub. Judges said no real world application. 3,100 forks later. The mathematician had been one of them. He had wired the agent into his content pipeline the week Scale AI cut him off. He had been a PhD candidate at one of the top five Chinese math schools. He taught there for two years before going full time on Scale AI contracts. He still has the credentials. He still has the office. He just no longer writes anything. He wanted to show people how AI took his career. He accidentally showed them how AI also took his post mortem.

11,906 次观看 • 1 个月前

A Chinese developer in Shenzhen was scrolling Twitter when he saw a 31 second clip of a Tokyo University student saying the trick to anything is starting for 5 minutes. He opened Claude and typed one line. Build me something that changes my life. You have 5 minutes. The agent did not ask follow up questions. It started writing code. By the time the 5 minutes were up Claude had built itself a second agent. The second agent's only job was to spawn a third one. The third one was watching Bitcoin. By morning there were 11 agents on his laptop. Each one running on the same one line prompt. Build something that changes his life. You have 5 minutes. Some agents wrote scripts. Some agents wrote other agents. One of them went straight to a public market and started entering positions. While the West is still asking Claude to write essays, China is already letting one prompt spawn 11 agents that change a developer's life overnight. The developer wanted to test the Tokyo student's idea. He just showed too much. He filmed himself walking through what the agents had built. Posted the breakdown to Bilibili. Pause at 0:21. Ignore his face. Look at the laptop on the desk. That tab is not the agent dashboard. That is a live wallet. kingofcoinflips. $820,819 profit. 3,189 predictions. Joined August 2025. → The Tokyo student was telling viewers how to start studying. One of the 11 agents had been running on its own for 8 months. Started with rent money the developer could not afford to lose. Now sitting at $820K. The agent does the simplest thing possible. Waits for Bitcoin to already make a big move on Binance. Then enters the side that already won while the public market is still catching up. $44K bet returned $114K. $9K bet returned $42K at 330%. Entry at 23 cents. Exit at $1. Biggest single win: $69,849. From a Tuesday morning Bitcoin window. The comment section turned into a detective board. Someone slowed the video to 0.25x. Read the wallet name off the corner of the laptop. Ran it through a public explorer and posted the profit curve. Flat for two weeks. Then straight up. 3,189 winning bets in a row. Every line green. The Tokyo student told viewers to start for 5 minutes. The developer typed one sentence and let the agent decide what to start. 8 months later the laptop is still deciding. He deleted the laptop frame from the reupload. Too late. Someone had already screen recorded it. The clip hit Discord. Then Telegram. Then every dev forum. The original tutorial got 2 million views. The zoom on the laptop got another 600,000. 99K people watching the wallet now. The student is still telling Japan to study for 5 minutes. The 11 agents are still rewriting their own code. The one in the corner is the one that actually changed his life.

A Chinese developer in Shenzhen was scrolling Twitter when he saw a 31 second clip of a Tokyo University student saying the trick to anything is starting for 5 minutes. He opened Claude and typed one line. Build me something that changes my life. You have 5 minutes. The agent did not ask follow up questions. It started writing code. By the time the 5 minutes were up Claude had built itself a second agent. The second agent's only job was to spawn a third one. The third one was watching Bitcoin. By morning there were 11 agents on his laptop. Each one running on the same one line prompt. Build something that changes his life. You have 5 minutes. Some agents wrote scripts. Some agents wrote other agents. One of them went straight to a public market and started entering positions. While the West is still asking Claude to write essays, China is already letting one prompt spawn 11 agents that change a developer's life overnight. The developer wanted to test the Tokyo student's idea. He just showed too much. He filmed himself walking through what the agents had built. Posted the breakdown to Bilibili. Pause at 0:21. Ignore his face. Look at the laptop on the desk. That tab is not the agent dashboard. That is a live wallet. kingofcoinflips. $820,819 profit. 3,189 predictions. Joined August 2025. → The Tokyo student was telling viewers how to start studying. One of the 11 agents had been running on its own for 8 months. Started with rent money the developer could not afford to lose. Now sitting at $820K. The agent does the simplest thing possible. Waits for Bitcoin to already make a big move on Binance. Then enters the side that already won while the public market is still catching up. $44K bet returned $114K. $9K bet returned $42K at 330%. Entry at 23 cents. Exit at $1. Biggest single win: $69,849. From a Tuesday morning Bitcoin window. The comment section turned into a detective board. Someone slowed the video to 0.25x. Read the wallet name off the corner of the laptop. Ran it through a public explorer and posted the profit curve. Flat for two weeks. Then straight up. 3,189 winning bets in a row. Every line green. The Tokyo student told viewers to start for 5 minutes. The developer typed one sentence and let the agent decide what to start. 8 months later the laptop is still deciding. He deleted the laptop frame from the reupload. Too late. Someone had already screen recorded it. The clip hit Discord. Then Telegram. Then every dev forum. The original tutorial got 2 million views. The zoom on the laptop got another 600,000. 99K people watching the wallet now. The student is still telling Japan to study for 5 minutes. The 11 agents are still rewriting their own code. The one in the corner is the one that actually changed his life.

17,952 次观看 • 2 个月前

Asked the owner of a $1.1M bot: How much did you spend before first profit? $47,000. Servers, developers, testing, money lost on bugs. People on Twitter say $100 and Claude Code. That's a lie. He broke it down: → $8,400 co location server in New Jersey (Equinix NY5, same datacenter as Polymarket nodes) → $12,000 freelance developer from Ukraine who actually understood latency → $6,200 three months of losing money while testing → $9,800 rewriting the whole thing when Polymarket changed their API → $10,600 bugs that bought at wrong prices, sold at wrong times, once bought both YES and NO on the same market → Profile: The $100 Claude Code posts get millions of views. Mine took $47K and six months. Nobody wants to hear that. 0x1979ae. 25,123 predictions. $1.13M profit. All BTC windows. Joined January 2026. I asked: Would you do it again? I made $1.1M. So yes. But I almost quit at month four. Balance was negative $23K. My wife thought I was gambling. The repo that finally worked: 3,400 lines. Not 47 lines. Not 23 lines. Three thousand four hundred. His one advice: If someone says they built a profitable bot in a weekend they're either lying or they copied a wallet instead of building one. The real edge isn't Claude Code. It's six months of losing money until you stop losing. → Copy: 183K watching. $0 loaded right now. He extracted everything. The bot is paused. Waiting for the next cycle. Twitter shows you the winners. Kreo shows you what it actually costs.

12,225 次观看 • 4 个月前

没有更多内容可加载