Topic Thread

#ai-evaluation

2 published observations tracking this thread.

AI forecasting probabilities being compared with real-world outcomes

Mantic's Forecasting Win Turns AI Prediction Into a Measurable Product

Forecasting is one of the rare AI applications where success can be scored against future reality, making Mantic's tournament result more interesting than a generic claim of better reasoning.

2026-09-19#Mantic · #forecasting

Android Bench 2.0 measuring long-horizon AI coding tasks, agent workflows and continuous completion scores

AI Evaluation & Developer Systems6 min read

Android Bench 2.0 Changes the Question From Code Completion to End-to-End Engineering

Android Bench 2.0 is a useful signal because its hardest tasks look more like real engineering projects than isolated coding exercises, and its low full-task pass rate makes that difference measurable.

2026-09-19#Android Bench · #AI coding agents