What it does
Jev estimates whether covered squares are safe from visible clues, and the harness clicks its safest pick. The maker’s animation is a replay of recorded games, not live inference.
Playable demos & bots
Watch Jev and a fly-connectome player race across separate Minesweeper boards, one move at a time.
Only you can see your notes.
Recording unavailable. Open the experiment ↗
Jev estimates whether covered squares are safe from visible clues, and the harness clicks its safest pick. The maker’s animation is a replay of recorded games, not live inference.
Sketch a small grid game where some hidden tiles hold hazards. Note what numbers or hints nearby tiles show when uncovered. Have a developer review the project's Python code to see how it sends visible tile clues to an AI to estimate which hidden space is least risky.
Your developer can then build a routine that automatically reveals the space the AI considers safest. Remember that AI judgments about hidden hazards are only educated guesses, so the program can still pick a dangerous tile by mistake.