I Gave GPT 6 Astra $10,000 to Trade Stocks...And This Happened
Published
Should I watch this?
Skip to the real contentYour full summary is ready below. Read on while we finish your watch recommendation.
Your summary is ready. The watch recommendation is next.
Worth watchingSkim
Skim the scheduled-agent setup, email-to-broker handoff, and final $100 loss versus the S&P’s $20 loss; most of the rest repeats balance updates and requests for more risk without explaining trade selection or demonstrating a reliable trading edge.
The main answer
The AI-assisted trading experiment lost to the S&P 500. After seven trading days, the creator reported turning $10,000 into about $9,900—a $100 loss, or 1%. The same investment in the S&P would reportedly have ended at $9,980, down 0.2%. That means the experiment underperformed by $80, or 0.8 percentage points—not the transcript’s stated “8%.” It did not demonstrate a profitable or superior trading strategy.
Key takeaways
- This was a human-adjusted AI experiment, not a hands-off test. The creator repeatedly pushed for more activity, loosened risk filters, and increased the proportion invested.
- When the main model refused automatic trade execution, the creator introduced a separate bot to execute its recommendations through Alpaca.
- The portfolio initially stayed mostly in cash, then became almost fully invested. More activity produced larger fluctuations, but not an overall win.
- Beating the S&P on individual days did not offset underperformance across the complete challenge.
- Seven days, changing rules, and one account cannot establish a reliable investing advantage. Some reported daily changes also do not reconcile cleanly with the stated balances.
- More aggressive trading does not necessarily improve returns. Buying dips does not guarantee recovery, and options can magnify losses as well as gains.
Summary
The creator documented a seven-day, $10,000 stock-trading challenge using AI research, scheduled reviews, and bot-executed trades. After repeatedly making the strategy more aggressive, the account finished below both its starting balance and the benchmark. He acknowledged losing, promised the giveaway, and proposed continuing the experiment.
What’s being promoted?
The clearest promotion is for the creator’s AIS Live event on October 17–18 and his free community. The consolation prize is a VIP ticket, and eligibility requires commenting on a post inside that community. This connects the trading challenge directly to community growth and event awareness. The video also showcases tools named as “Codeex,” “GBT6 Astra,” “Grockbot,” Alpaca, and Open Claw; some names may be transcription errors. Calling setup simple and demonstrating the workflow gives these tools favorable exposure, but the transcript does not establish sponsorship, affiliate payments, or other commercial relationships. An additional apparent goal is entertainment and continued audience engagement. The creator says he would have preferred losing $500 to seeing little activity, repeatedly requests greater aggression, asks viewers about longer challenges, and promises follow-up videos. That suggests dramatic movement and recurring content matter alongside investment performance. His closing emphasis on the first two days being too passive also softens the failure and preserves interest in another attempt. These incentives do not prove dishonesty: he openly reports the loss. However, the transcript alone cannot verify the real-money claim, complete trading record, giveaway fulfillment, or undisclosed business relationships.
