Simulate a choice forward before deciding

Writes the question, options and measure of better, labels every uncertain input as measured or guessed, locks the model and seed and runs every option across many runs. For a hard choice under uncertainty.

What it asks for

  • about (Optional, words): What this run is for, in the person's words.
  • known (Optional, words): What is already known, decided or done, so the run starts from it instead of asking again.

The steps your AI walks

  1. Say what is being chosen, the options, what counts as better and the model
  2. Gather every uncertain input as a labelled entry, measured first and guessed last
  3. Lock the inputs, the model and the seed before any result exists
  4. Run every option forward across the runs and the steps
  5. Write what each option did as result entries
  6. Find which inputs flip the winner
  7. Say what wins, on what it rests, and the cheapest real test of the riskiest guess
  8. When reality answers, keep it, rerun and record what matters most

Use it from your AI

Connect the AI you already use to Trillion once. Then ask in your own words; your AI finds this workflow by what it does and runs it.

Say to your AI: “Use Trillion to writes the question, options and measure of better, labels every uncertain input as measured or guessed, locks the model and seed and runs every option across many runs. For a hard choice under uncertainty”