Arena lands $200M to measure what AI agents actually do
The Series B values the AI evaluation company at $3.1B as agents…
Vals wants to grade AI on real work, not exams
A two-year-old benchmarking startup backed by Andreessen Horowitz keeps its test sets…
Why measuring AI progress is getting harder
https://www.youtube.com/watch?v=WO9c9qxDxzU Public benchmarks are saturating and models are getting better at optimising…
