Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

it is far less willing to provide code, I get better results and faster out of 3.5.

I am at the point that 4.0 is basically not worth using as single entity, but it seems that using the api and generating some combative/consultative agents yields some interesting results, but not super fast.

Check this out if you have not seen it already : "AutoGPT Test and My AI Agents Effortless Programming - INSANE Progress!"

https://www.youtube.com/watch?v=L6tU0bnMsh8



Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: