I agree with this. I’m a paid customer of both AIs. I also use image generation and voice chat quite a bit so keep ChatGPT for these features but I go to Claud first for everything else. I like how Claude will say it doesn’t have enough knowledge to answer some specific questions while Chat GPT just makes it up.
"Wow, SmythOS's agent templates sound like a game-changer! The idea of AI agents collaborating like human teams is mind-blowing. The potential seems limitless! 🚀"
We'll see. Anthropic is still going to release Claude 3.5 Opus in a few months and if it's as much better as Claude 3.5 Sonnet as Claude 3 Opus was to Claude 3 Sonnet, it'll be an insanely strong model. However, the stuff with Q* and strawberry does make it sound like there could be something coming up which is truly qualitatively different from anything else we've seen thus far. My intuition is that Claude 3 Opus will be similarly capable as GPT-5 (or however they call it). I think Q* is too computationally expensive to be deployed widely if the rumours about how it works are roughly accurate.
Great examples! Is there a way to "store" the things I have created with claude somewhere? For example, how would I do this for flash cards? Because coming back to the chats again seems tiring
Is there any restrictions to Claude? I am a pro user since the beginning, but I cant see the terminal where the code is being tested as you showed for the all the examples
