I test AI against my own investment decisions. This is what I find.
A personal record of AI experiments run against real money — what worked, what didn't, and what I learned.
Why I now split every serious AI question into stages before I trust the answer
How I use a four-stage prompting method — Role, Filter, Risk, Verdict — to get useful AI analysis instead of confident-sounding noise.
Five things AI is reliably bad at for investing
The failure modes I've hit often enough to warn you about. Each one came from a real position, not a hypothetical.
Lab ReportI asked AI to find the downside in a position I was already long. It found things I'd missed.
An adversarial prompt against my own portfolio. The model surfaced three risks I'd glossed over, and one I'd actively been wrong about.
Lab ReportI gave the same investment question to four AI tools. The results were instructive.
Not a ranking. A controlled test on a real question, with notes on what each tool actually produced and why most of it was useless.
A more rigorous way to use AI.
The Prompt Stack is how I turn a general model into a narrower analyst. Instead of asking for instant brilliance, I break the task into stages, force evidence before opinion, and make the final output earn its keep.
Get the Lab Book — it's free.
Prompt Stack templates, decision checklists, and experiment notes. Free when you subscribe.
Ben Dixon documents AI experiments against his own portfolio. Real money, human analysis, sceptical use. Read more →