I think another reason OpenAI talks about how scary LLMs are is that they fear the open source models surpassing them. I didn’t see it mentioned in the article, but both Google and OpenAI have called for regulation and I think that they only want that because they have no moat.
Super interesting perspective! I have a pretty non technical friend that got pretty into stable diffusion a while back and was able to do really cool things with it. The idea that LLMs can be that accessible is really exciting.
I think another reason OpenAI talks about how scary LLMs are is that they fear the open source models surpassing them. I didn’t see it mentioned in the article, but both Google and OpenAI have called for regulation and I think that they only want that because they have no moat.
Super interesting perspective! I have a pretty non technical friend that got pretty into stable diffusion a while back and was able to do really cool things with it. The idea that LLMs can be that accessible is really exciting.
Regulations would just mean that only a corporation could find enough training data to use.
Fortunately, and country to ban and restrict risks being outcompeted. Hopefully this discourages them.