My AI Trading Bots Lost $130... So I Hired an AI
Creator Magic
0:00 / 0:00
My AI Trading Bots Lost $130... So I Hired an AI
12 074 просмотра · 19 часов назад
Creator Magic
223 тыс. подписчиков
12 074 просмотра · 19 часов назад
🐝 Copy my Beekeeper ⚡️ free template: https://mrc.fm/beekeeper
My 3 AI trading bots lost $130 of their $1,000 in one week, so I built the Beekeeper in Zapier: Jev decides which bot is broken or just unlucky, and Claude Opus 5.5 rewrites its trading rules. A self-improving AI trading bot loop, in one Zap.
Watch the bots trade crypto live with real money: https://beebots.tech
Build your own AI trading bots, free and open source (MIT licence): https://github.com/imikerussell/beebots
Part 1, I Gave 3 AI Trading Bots $1,000 (Jev): • I Gave 3 AI Trading Bots $1,000 (Jev)
WHAT YOU WILL SEE
Three AI crypto trading bots (Bizzy, Breezy and Boozy) one week into trading real money, and why a decision model like Jev can referee every trade but cannot fix a bad strategy. Then the build: one Zap, triggered by a webhook every hour, reads the scorecard, asks Jev whether an AI trading agent is broken or just unlucky, and hands the broken one to Claude Opus 5.5 in AI by Zapier to write new trading rules. Every decision is logged in Zapier Tables, a Buzz agent builds part of the Zap with computer use, and the new rules go live on beebots.tech with a one liner from the Beekeeper.
NOT FINANCIAL ADVICE. This video is for education and entertainment only. Trading crypto and other assets is risky and you can lose all of your money. Crypto can go to zero. The bots run on paper trading by default and paper trading is advised. Nothing here is a recommendation to buy or sell anything. Do your own research.
This video is sponsored by Zapier.
CHAPTERS
0:00 3 AI trading bots, $1,000, down $130
1:23 AI trading bot results after 1 week (live dashboard)
2:17 Jev made 600,000 trading decisions for $10
2:34 93 viewers built their own AI trading bots
2:58 Open source trading bot rules you can copy
3:23 Why AI trading bots lose money: a referee, not a coach
3:59 The Beekeeper: a self-improving trading bot in one Zap
4:29 Scorecard step, then Jev decides: broken or unlucky?
4:48 How to use Jev in Zapier: API key and the TypeSafe Jev app
5:28 What data to send Jev
6:35 Jev picks the failing trading bot
7:04 Handing off to a frontier model
7:19 Zapier filter step: only continue if a bot is named
7:45 Claude Opus 5.5 inside AI by Zapier
8:52 Claude Opus 5.5 rewrites the trading rules
9:21 Zapier Tables: logging every decision
9:49 An AI agent builds the Zap with computer use (Buzz)
11:17 The Beekeeper goes live on beebots.tech
11:31 How the self-improving loop works in 20 seconds
11:55 What it costs to run
12:39 Zapier pricing and how to build your own