Y
Hacker News
new
|
ask
|
show
|
jobs
by
dev_l1x_be
61 days ago
Benchmarking is grossly misleading. Claude’s subscription with Code would not score this high on the benchmarks because how they lobotomized agentic coding.