You missed what they meant. It means gpt5 is really good at one arbitrary and extremely specific topic. Anything else it’s comparable with a random person on the street.
Reality is the opposite though. GPT5 is expert in a pretty wide amount of trivia. It’s better than the average uneducated person in every subject, but worse than an expert in every subject.
All the AI are wrong at least 1 in 20 tokens, probably more like 3 in 20, to their sample data which is also wrong a considerable amount of the time, so it is an expert in literally nothing and never will be.
That 1 in 20 number comes from the 2020 paper by OpenAI and the 2022 paper correcting it by Deepmind about AI scaling laws where the AI inverse error rate would never reach 94% with infnite training and power.
You missed what they meant. It means gpt5 is really good at one arbitrary and extremely specific topic. Anything else it’s comparable with a random person on the street.
I’d say we’re actually worse than the average person at everything else. Too much of our brain is allocated to our research.
It does seem like there’s an inverse correlation of general intelligence/common sense and specialized study.
Reality is the opposite though. GPT5 is expert in a pretty wide amount of trivia. It’s better than the average uneducated person in every subject, but worse than an expert in every subject.
And the average person has (usually) no idea if what they’re being fed is even correct. But as long as it sounds correct to someone who has no idea…
Plausible bullshit machines.
That’s my favorite Nine Inch Nails album!
All the AI are wrong at least 1 in 20 tokens, probably more like 3 in 20, to their sample data which is also wrong a considerable amount of the time, so it is an expert in literally nothing and never will be.
That 1 in 20 number comes from the 2020 paper by OpenAI and the 2022 paper correcting it by Deepmind about AI scaling laws where the AI inverse error rate would never reach 94% with infnite training and power.
deleted by creator
GIGO
It can also have good in and garbage out, as I mentioned it will never achieve human input accuracy much less superhuman input.