The Harness · Today · 4 min
Same-score coding agents split once the tasks change — Nov 8
0:00-4:43
transcript
show notes
MaxxBench perturbed 89 DeepSWE coding tasks and turned a six-way tie near 74% into a 23-point spread, raising fresh doubts about public coding scores. Claude Code's saved approvals have put secrets into more than a thousand public GitHub files, while a developer's switch to the Pi harness points at prompt churn as a lock-in weak spot. Microsoft Research's AI Frontiers group has been folded into Microsoft AI, and reverse engineers say Apple Intelligence ships an undocumented family classifier.