Lighting the engine…
Flint collides a resource you'd recognise with a move you'd never put beside it, reads the business the collision throws, and grades it honestly. No blank page, no prompts to engineer — the same engine for your first idea or your hundredth.
What a strike looks like
Real engine output — the first strike in the demo above, one stage at a time.
Everything above is verbatim recorded output. From the verdict you keep going — drill, fuse, restructure, regrade — each step a control you press, judged again after every move.
The method
A vocabulary of resources and moves, an engine that pairs them for surprise, a hunt that evolves the idea one graded step at a time.
The engine picks coordinates deliberately far apart — distance is where surprise lives.
Every strike lands broad, focused, and specific. Pick an altitude; the interpreter reads the business.
Drill, fuse a second resource, restructure, regrade against a market — re-judged after every move.
How it works
Every reading ends in a grade — judged by a grader that is walled off from the interpreter. It never sees how the idea was built, only the finished idea, so its verdict isn't quietly talking its own book.
Real, untaken edge — worth chasing.
Sound but well-worn — the surprise has been used up.
No edge at all — everyone who feels this is already served.
The grade measures surprise and untaken edge, not profit — a profitable but familiar idea grades tame on purpose. The demo's second strike is an honest tame: fuse a second resource into it and watch.
What Flint won't show you is how it chooses the collision. That part is the engine. And a plain word on the obvious: readings are AI-generated — treat a flicker as a lead worth investigating, not vetted business advice.
Four ways in
Wherever you begin, Flint ends the same way — with a graded idea and an honest verdict.
An idea that doesn't surprise is one the market has already answered. Flint is built around that single conviction. It doesn't reward ideas for being sensible or lucrative — plenty of tame ideas are both. It rewards them for carrying edge no one has taken yet, and it's honest when they don't.
Pricing
No subscription. You buy credits; every action spends in proportion to the work it does, with your balance always on screen. A typical hunt — strike to graded idea — runs about a dollar.
Room to drill, fuse, and regrade without watching the meter.
Buy 33 creditsCredits are valid for 12 months from purchase. Launch pricing — pack sizes and rates may be tuned as the engine's costs settle.
Three recorded strikes are free above. The fourth one is yours — your resources, your pains, your verdicts.
Start for $1