Developer hands an LLM $10k to trade: X signals + filings drive an autonomous portfolio
demian_ai · x · 2026-09-26
Developer demianai is running an experiment letting an LLM autonomously manage a $10,000 portfolio, asking it to run a portfolio rather than pick stocks.
The stack combines:
- X signals and web search (hand-picked top voices plus automated search)
- Perplexity research and Tavily retrieval
- Company filings, earnings, analyst targets, news, technicals and catalysts, private reports
Scope is limited to 223 stocks in AI infra, energy and robotics. Each session, MiniMax M3 (running via Nebius) decides what to buy, hold, add, trim or avoid, producing a target weight and multi-horizon forecasts for every stock.
Related event: Developer gives LLM $10,000 to autonomously manage a portfolio(3 posts)→
More from coding & agent
- OpenRouter open-sources Jev Router, a cache-aware router that picks the best model per request — markjeffrey · 2026-09-26
- 4B model mines an iron pickaxe in Minecraft in 23 decisions, generating zero tokens — Top-Evidence174 · 2026-09-26
- Open-source Jev router cuts LLM token costs ~9x by routing Claude Code and Codex traffic — airesearch12 · 2026-09-26
- Mica v0.1 4B: Open Decision Model Runs on 8GB GPU, Trained for Under $30 — Top-Evidence174 · 2026-09-26
- Solo builder wins $500,000 Gemini XPRIZE beating 26,000 entrants with agent-built Polyfork — Scobleizer · 2026-09-26
- NetHack run used a self-built harness with moves logged on a remote server — gleech · 2026-09-26