TRACES Benchmark Launched to Evaluate 'Discoverative AI' Beyond Known Answers
Aiden_Tech_Ai · x · 2026-08-22
Apodex has released TRACES, the first benchmark designed for "Discoverative AI"—systems that can solve problems without existing answer keys.
Unlike traditional benchmarks that test retrieval, TRACES evaluates the entire investigative journey, including evidence processing, self-correction, hypothesis testing, and establishing conclusion boundaries. It features scientific tasks like AAV capsid design, drug repurposing, and clinical trials.
Proposed by founder Tianqiao Chen, the release includes a definition of discoverative intelligence, a rubric for distinguishing valid investigation from luck, and an open call for solvers and new problems.
Related event: Apodex Launches TRACES, First Benchmark for Discoverative AI(8 posts)→
More from AGI Musings
- Current AIs share our fate: misaligned ASI would doom them too — louisvarge · 2026-10-16
- Testing alongside an agent in a p2p room feels oddly real, says dev — sull · 2026-10-16
- Yudkowsky's 2008 takeoff scenarios resurface as AI-advances-AI takes hold — jd_pressman · 2026-10-16
- Guardrails can make risky delegation work — but it takes a full-time team, says iandanforth — iandanforth · 2026-10-16
- e/acc founder: agentic models are autocomplete for digital actions — beffjezos · 2026-10-16
- What can you truly delegate? Only tasks with unambiguous, verifiable results — iandanforth · 2026-10-16