16:14
2026-09-29
dev.to
ai-agents
We recompute TypeSafe's 444x claim β here's what we found
A developer at Assay (δΌζ΄η§ζζιε ¬εΈ) built a 30-question exam to test whether AI agents can recall their own operational history, and found their own grader was wrong on 7 of 18 items, including two incorrβ¦