AI Matches Human Experts in Key Fields

AI is now officially on par with human experts in a ton of different jobs. OpenAI just dropped

this new benchmark called GDPval that tests how AI performs against actual professionals in fields like finance,

healthcare, and software engineering. They had humans compare reports written by other

pros to reports written by AI. Get this: Anthropic’s latest model was rated as good as

or better than a human expert in nearly half the tasks. OpenAI’s new GPT-5

wasn't far behind, scoring triple what the last model did just 15 months ago. The progress is

insane. For now, the test is pretty limited to just writing reports.

But OpenAI says this isn't about replacing people. It’s about letting AI handle the grunt