Judgment Gym
Practise the calls AI can't make for you.
Judgment Gym is a practice gym for people making consequential decisions with AI-generated work. Each case hands you imperfect evidence. You inspect it, commit to a call, get useful feedback, and update when the facts change.
AI can draft the memo. It can't tell you whether to trust it.
More of the evidence in front of a founder is now written by a model: summaries, forecasts, dashboards, drafts. The skill that matters is spotting what is wrong with it and making the call anyway.
That skill is rarely practised on purpose. Judgment Gym gives it reps.
How it works
Inspect
Read a realistic artifact: a memo, a dashboard, an AI-written draft, with the flaw sitting where a skimmer would miss it.
Commit
Make the call. The starter path needs no typing at all.
Get feedback
See what your path noticed and what it missed, with feedback written for that choice.
Update
The facts change. Decide whether your call still stands.
Inside the gym
What makes it different
No single right answer
A different defensible call, argued well, does as well as yours. The feedback judges the reasoning, not the verdict.
Your words as evidence
In written practice, feedback quotes your own sentences back to you on five fixed dimensions.
Private by default
Your responses are yours. Sharing is opt-in.
Where it stands
Beta, as of October 2026.
Built
- A working prototype with 12 scenarios, seven chapters and a first briefing.
- A practitioner-reviewed first case family.
Now
- The choice-first player, so a case can be played without typing.
Next
- Testing with learners, then the free public beta.