You must be American. I am talking about Kimi, Mistral, GLM, Liquid, Minimax, Arcee, Qwen, Deepseek, Xiaomi.
And you are of course allowed to use cloud inference if you don’t have the hardware to run locally. Just choose an inference service that is not in bed with fascists. There are plenty. Good luck and have a nice day.
I only recognize the last two of those. Which fascists are you referring to? There’s too many in the world right now to keep track. Since we’re talking about Claude being the least evil option, I’m not sure how those other companies tie in to the conversation.
Which would you recommend? I’m curious where your recommendation fall in the rankings of usefulness.
The companies I listed, as far as I can tell, are not supplying or trying to supply any fascist military. Some (like Alibaba who made Qwen) are obviously unethical in other ways. But all these companies make open-weight models so if they are unethical you can always avoid contributing to them by selfhosting or going to third-party inference providers.
Kimi is my favorite. But you can try GLM going by your leaderboard website.
You must be American. I am talking about Kimi, Mistral, GLM, Liquid, Minimax, Arcee, Qwen, Deepseek, Xiaomi.
And you are of course allowed to use cloud inference if you don’t have the hardware to run locally. Just choose an inference service that is not in bed with fascists. There are plenty. Good luck and have a nice day.
Sadly.
I only recognize the last two of those. Which fascists are you referring to? There’s too many in the world right now to keep track. Since we’re talking about Claude being the least evil option, I’m not sure how those other companies tie in to the conversation.
Which would you recommend? I’m curious where your recommendation fall in the rankings of usefulness.
The companies I listed, as far as I can tell, are not supplying or trying to supply any fascist military. Some (like Alibaba who made Qwen) are obviously unethical in other ways. But all these companies make open-weight models so if they are unethical you can always avoid contributing to them by selfhosting or going to third-party inference providers.
Kimi is my favorite. But you can try GLM going by your leaderboard website.