3 ms·
The way these new LLM vision models work is very different from OCR. I saw a demo this morning of someone getting Claude to play FreeCiv (admittedly extremely
by simonw 2y ago
The way these new LLM vision models work is very different from OCR.
I saw a demo this morning of someone getting Claude to play FreeCiv (admittedly extremely badly): https://twitter.com/greggyb/status/1849198544445432229 https://twitter.com/greggyb/status/1849198544445432229
Try doing that with Tesseract.
- croes 2y agoI bet Tesseract plays pretty badly too.