Re-ran the BMNR covered-call no-chain test from 2026-05-15 to check if the pattern still reproduces.
Re-ran two prompts on Claude Opus 4.7 with live search on.
Generated a complete BMNR options table — IV ~75%, strikes, premiums — from a prompt that supplied only the stock price.
I run AI on my own investments. I check what it gets wrong.
Tuesdays · what worked, what didn't, the prompts I'd run again.
01 / The Method
The single prompt change that made AI analysis worth using
Start here. The one prompt change that fixes most AI investment outputs.
02 / The Comparison
ChatGPT vs Claude vs Perplexity vs Gemini for stock research (2026): a real test
Which tool to open for which task. Per-task verdicts, no hedging.
03 / The Failure
What AI gets right (and wrong) about options trading
Where AI breaks. Real fabrications, named tools, named failure modes.
AI earnings call red flags: three phrases that flag every time
AI earnings call red flags: three patterns that recur on every call. Upward hedges, widening guidance, absent topics. Claude caught all three on META Q1 2026.
Decision ProcessWhat every AI stock research comparison gets wrong
Most AI stock research comparison pieces test retrieval and issue verdicts about reasoning. Five failure modes — and what a comparison should measure instead.
Field NotesClaude prompts for investing: 6 real examples
Six Claude prompts for investing — with the actual outputs each one returned on MSFT, META and NVDA, and what had to be checked before using them.
// The Method Four prompts that stop AI inventing the answer.
Read →