Skip to content

I build AI agents for my real portfolio. Here's what actually works.

A personal lab journal of AI workflows, prompts, and agents tested against real portfolio decisions.

Real-money testing
Human-written notes
Sceptical AI use
// Decision Process Featured · Latest
· 5 min read

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.

Portrait of Ben Dixon
Ben Dixon
Test Pilot
// The Prompt Stack

Stop Asking Lazy Questions.

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.

01ROLE: Review the position as a cautious analyst, not a cheerleader.
02FILTER: Separate observable facts from assumptions or model filler.
03RISK: Surface timing, downside, and what would invalidate the view.
04VERDICT: Give one practical action, then state the confidence level.
// The Lab Book

Get the actual working material.

Prompt Stack templates, decision checklists, and notes from live experiments. Join the list for new lab reports and Lab Book access.

Ben Dixon
// About Ben

Ben Dixon documents AI experiments against his own portfolio. Real money, human analysis, sceptical use. Read more →