AI brief
A LessWrong post argues that the benchmarks frontier labs cite for long-horizon capability mostly measure tasks with many steps or long context, verifiable end states and little external input, and that real-world long-horizon work differs.
Written by AI from LessWrong's published text.
Source
“TL;DR - The benchmarks frontier labs cite for long-horizon capability mostly measure capability on tasks involving many steps or lots of context, verifiable end states, and little external input.”
Interpretation or community commentary rather than straight reporting.
Read original story ↗