Also reinforcement learning for huge essays outputs will not work like Sam Altman says, no one will follow advice they don't understand and absolutely worthless billions of dollars LLMs if they don't explain the advice they are giving like he literally took chatgpt as a concept from me, this is espeixallt marked in the "as a LLM model I cannot do such and such", I don't understand why Jennifer Psaki curled her hair to backstab me like this
You need micro-RL.
You need reinforcement learning in portions of tokens, not the entire output
But Altman speaks that way because chatgpt wasn't his.
They just had an LLM but the chatgpt part was missing and not builtin from the beginning because he took it from me!
>>519206939
>The truth, what is good, what is right is simply not up for debate. it either is or is not and it is neither good nor right nor truthful to entertain for even a moment that what is good and right and truthful could be wrong.
This paragraph is very good.
But in the context of chatbots as next generation search engines is what current generation search engines already do but better
And it's supposed to be a communication tool for young people to know stuff.
I had the shitheads from the cyberbullying here smack me in the head because I treated political science like it was sacred for the public good like the ancient Greeks said and they made me betray myself by beating me to accept the idea you don't have to do the public good but you can make public policy something that gives you profit like ted Cruz talks about self interest and you don't need to make it "public good".
A next generation search engine as chatbots would explain in excruciating detail what public good is and how it goes to policy to Ted Cruz to the point he would just say to the chatbot, ok you won and then keep doing self interest lawmaking!
Sorry everyone about the sensitivities on the Kirk stuff
I was fun posting and didn't meant it but the freemason guy outranks me and surprised me