These models are mad libs machines. They just decide on the next word based on input and training. As such, there isn’t a solution to stopping hallucinations.
These models are mad libs machines. They just decide on the next word based on input and training. As such, there isn’t a solution to stopping hallucinations.
Proton is just a company and would always give in to legal requests and all other companies and any email provider would do the same
It’s amazing how people easily forget about lavabit and what a company that is committed to real privacy is about.
Let’s hope they move to Paizo and join the work on pathfinder 2.5.
This is going to be like the self checkout lanes at the store but for creative jobs.
At the end of the day, a company will be able to produce the same output with fewer people. Some stuff will be of lower quality, just like sometimes people spend time on Lemmy and then phone in some crappy work.
We’ll, strictly speaking you could have an AI that only knows about the world up to 1928 and talks like it’s 1928.
Of course. He is only a chatting predator or pedofile.
/s btw. I don’t know anything about the situation