3 ms·
If LLMs are so great at learning by themselves, why does OpenAI need to resort to the plug-in model? Is it because it can’t actually do logic (hence the Wolfram
by andsoitis 4y ago
If LLMs are so great at learning by themselves, why does OpenAI need to resort to the plug-in model? Is it because it can’t actually do logic (hence the Wolfram Alpha plugin)? Or is it because that’s the way it gets access to more data? Or both?
https://openai.com/blog/chatgpt-plugins https://openai.com/blog/chatgpt-plugins