Interactive calculator

Exploring AI: Data Napkin Math

Interactive napkin math, back-of-the-envelope estimates, and ballpark figures for training-data value, compensation, and distribution questions. How will the proceeds and benefits of AI be distributed?

0 curated scenarios shown, plus one open-ended builder at the end. Inputs are shared across cards, so one edit updates every calculation that uses that variable. Within each card, assumptions are grouped by question type.

Scenarios

Change the assumptions directly on each card. Inline comparison menus let you swap in related public benchmarks without leaving the page, and every input card links to its cited source.

1. Edit a number on any card.2. Swap a benchmark in the pill-shaped comparison chips.3. Share the exact state once the answer feels useful.

The inline chips inside each scenario description are live benchmark swaps, not just labels.

Distributing money

Build Your Own Scenario

Build a custom formula from the shared input library when the curated scenarios do not quite match the question you want to ask.

Custom result
Check formula
dollars

Enter a formula to start exploring.

Keep this advanced builder tucked away until the curated scenarios stop being enough.

Open the builder when you want a custom formula, markdown scaffold, and shareable draft.
About this page

This website is still evolving. The goal is to make public debates about training data legible through transparent assumptions and easy-to-share napkin math.

The calculations are intentionally simple. What matters most is helping readers inspect the assumptions, compare alternate reference points, and see how much each estimate depends on a small set of shared inputs.

To participate in the discussion about better defaults or additional scenarios, visit the GitHub repository.