AgentsCodingResearch

AI coding agents still need human cleanup on nearly three out of four tasks in a real-world test

via The Decoder · 35 views · 1w ago

AI coders aren't as smart as you think. A new benchmark shows even top AI coding agents, like GPT-6 Astra, only fully solve 26.4% of real-world tasks. That means humans still clean up almost three out of four problems.

Read the original story →

← Back to the feed

Loading comments…

Log in to join the conversation.