3 ms·
It was interesting to see GPT4 fail at it: https://chat.openai.com/share/02c12bbe-43cd-40da-b5df-33681db370d7 https://chat.openai.com/share/02c12bbe-43cd-40da-b
by CamperBob2 3y ago
It was interesting to see GPT4 fail at it: https://chat.openai.com/share/02c12bbe-43cd-40da-b5df-33681db370d7 https://chat.openai.com/share/02c12bbe-43cd-40da-b5df-33681d...
Not a bad benchmark problem. It didn't get very far, but maybe the next release will.
- lupire 3y agoChatGPT can't write programs or do logic to solve problems whole solutions aren't already in its input.
- CamperBob2 3y agoIt applied logical reasoning, just not quite enough of it. When it gets 10x better, which it will, it will be smarter than either of us. For those who don't find that prospect fascinating, other sites beckon.