Nate Herk | AI Automation
I Gave GPT 6 Astra $10,000 to Trade Stocks...And This Happened
Summary
AI summaries can be incomplete or wrong. Verify anything important against the original video.
Nate Herk attempts a 7-day stock trading challenge where an AI (GPT-6 Astra) manages a $10,000 portfolio to try to beat the S&P 500. Despite multiple adjustments to the AI's aggressiveness, the portfolio fails to beat the S&P 500 by the end of the week, resulting in a loss of approximately $100.
In this 7-day experiment, Nate Herk empowers an AI system named GPT-6 Astra, running on his 'Codex' platform, to manage $10,000 in real capital, with the goal of outperforming the S&P 500 index. Initially hampered by overly cautious, passive constraints that resulted in minimal activity, Herk iteratively adjusts the AI's settings, allowing it to make more frequent decisions and take more active positions in stocks. He introduces a 'Grok bot' to handle trade execution via email triggers once the AI identifies opportunities. Despite the increased activity, market volatility and the inherent difficulty of short-term stock trading lead the portfolio to underperform the index by the end of the week. Herk concludes that while this passive AI-managed approach is a learning experience, it requires significantly more sophisticated strategies than those used to achieve the desired profit-beating benchmark.
Key Points
LockedWorth watching if: You enjoy AI-driven experiments and want to see a practical, albeit unsuccessful, test of LLM-based autonomous stock trading. It is useful for understanding how constraints and prompt engineering affect the behavior of AI agents in real-world, high-stakes tasks.
Sign in to unlock the full extract
Every claim, key point, and timestamp for this Nate Herk | AI Automation video — plus a daily email of every channel you follow.
Sign in with GoogleNo credit card. Free tier forever.