AI Possibility-Space Mapping
Use AI to enumerate the full combinatorial space of a multi-dimensional decision and build intuition fast.
- Difficulty
- Easy
- Time to result
- ~days to results
- Steps
- 4
- Confidence
- 80%
A specific, underused way to use AI: when you face a space of possibility with many dimensions, have the model sample a subset of dimensions and lay out every combination (with titles and descriptions) so you can grasp the whole option space at once. Use it to build intuition when a decision has too many variables to reason through manually.
Origin
In the AI corner, Field shared a personal method that goes beyond everyday AI use: rather than asking AI for a single answer, he asks it to generate the full table of combinations across selected dimensions, which builds intuition about a possibility space in a way manual selection cannot.
Core principles
- 01AI is strongest not just at answers but at exposing a whole possibility space.
- 02Reduce an unmanageable space by sampling a subset of its dimensions.
- 03Enumerate every combination, each labeled and described, to see the field at once.
- 04The goal is intuition about the space, not one recommended output.
How to run it
- 1
Identify the dimensions
Name the many attributes or variables that define your decision space (e.g. a list of 100 personality traits for a fictional character).
- 2
Sample a subset
Have the AI randomly pick a manageable handful of dimensions (e.g. six of the 100) so the combinatorial space stays graspable.
- 3
Enumerate all combinations
Ask for the full table toggling each attribute positive and negative across every combination, with a title and description for each.
Pro tip Requesting a title plus description per cell makes the space navigable rather than an abstract matrix.
- 4
Read the space to build intuition
Study the completed table to develop intuition about the shape of the whole possibility space, not just to pick one option.
In the wild
Field described wanting to generate a character with ~100 possible personality traits. Instead of hand-picking, he has AI randomly select six traits and produce the full table of every positive/negative combination, each with a title and description.
→ A complete map of what that character sample space could look like, building intuition in a way manual selection can't.
Common mistakes
Asking AI only for a single answer
Requesting one output wastes AI's real strength at illuminating an entire multi-dimensional possibility space.
Trying to enumerate too many dimensions at once
Without sampling a subset, the combination table explodes and stops being graspable.
Is it for you?
Best for
Anyone facing a high-dimensional creative or strategic choice who wants to understand the option space, not just get an answer.
Not ideal for
Simple decisions with few variables where a direct answer is faster and enumeration is overkill.
From the transcript
“whenever you have a space of possibility and there are many dimensions to that space.”
“randomly pick six out of this list of 100 and then give me basically uh for every attribute uh the full table”
“It just builds intuition about a possibility space in a different way if you do that.”
From the episode
Figma’s CEO: Why AI makes design, craft, and quality the new moat for startups
Dylan Field