Skillprint Signal: AI Benchmark for Human Performance

Skillprint has launched the Signal Benchmark, the first AI benchmark grounded in real human performance. This new tool evaluates how effectively AI models can enhance real human performance. Here's the thing: existing benchmarks primarily measure what an AI model knows or how accurately it answers questions. Signal, however, measures how well a model adapts a live game in real time to move a real person toward a target mental state. This could be greater focus, relaxation, or improvement in a cognitive skill. Signal places AI models inside live gameplay with real participants. As someone plays, the model adjusts game parameters in real time to achieve a defined human outcome. Signal then measures the resulting change in the player's state and ranks models based on their effectiveness. This benchmark is built on six years of research involving over 500,000 players and a library of more than 1,000 games. At launch, Signal runs on two games, Hextris and Box Tower, and measures target states like Relax and Focus. Results will be published on a public leaderboard. The bottom line is this benchmark aims to measure if AI can truly help and enhance human capability, which is crucial as AI becomes more integrated into daily life.