AISilicon Canalsabout 3 hours ago
AI productivity studies show conflicting results
6 min read
Harvard and BCG found consultants using GPT-4 finished skilled tasks 25.1 per cent faster, but METR's trial found experienced coders using AI tools were 19 per cent slower. Developers still believed AI sped them up by 20 per cent despite the measured slowdown.
Level
Hype check
Tap to vote and see what everyone thinks.
Summary by ByteBrief
