For a while I have been thinking that all this nonsense about AI world domination and lousy safe guards on grok enabling people to do indecent stuff is just a way for fastforwarding us into passing regulations for hardware gated LLMs so that people can’t use open models too easily. After all, introducing bugs will shoot them in their own foot if open models surpass them.

As for shitty code, I think the more unmaintainable it is the better for them because then you rely on their services to maintain the code and eventually they might quite substantially ramp up prices provided open LLMs remain underdeveloped or hard to access.

Lol would be fun to watch all of big tech turning hostile to AI.