You don't think this is a response to corporate culture? Especially over the last few years.

Like just to start,how are you going to balance intelligence so that it's fun? The AI doesn't magically know.

AI is generally unpopular with people.

Think through an art career, some kid goes through high school, a fine arts degree or something. You get a job working on movies, video games something flashy. Now AI is going to take that All away.

Programming, maybe you start off making games or hacking,then you go too School and get a job. The job is a whole team or sr swes who don't care, your going through project planning ceremonies, playing politics, working on communication skills so you can survive in the corporate environment. You're just working on connecting some crm to some other API. You had to grind Leetcode problems to get that pointless job. What exactly are programmers protecting by hating ai? The job already sucks, it's already pointless, and the first world has been threatened by being replaced already by cheap 3rd world employees

Idk if people don't realize you can look this up? Pay doesn't look suppressed to me....

every time you ask for the next token you different gpus on different servers are used. each generation of a token is an independent operation from the previous.

a brain is a system of neurons, you can add or remove some, it still some how generates a conscious experience. an llm generating tokens is doing math in a distributed way across what ever gpu happens to be free for that isolated request. there is no consistent system. what do you consider to be the conscious entity? the random selection of gpus? you think something is conscious one token at a time? what is the exact thing that is thinking? the way you can point to a body, you cant do that for an llm string of tokens. thats where i think it falls apart. your doing math that eventually reduces into a final number. theres no one system that continues to run and take input in.

What's the system, the thing, that's experiencing consciousness when you talk about llms? Again, if you think through the request path of each token that's generated, this idea falls apart.

You need to think through the entire request for an llm model response end to end.

The same gpu can run multiple llm requests for different people.

My neurons don't handle experiences for other humans. It's not a multi tenant brain (I hope haha)

I'm pretty sure panpsychism has answers for most of the other questions. I'm sure a commentor will correct me though.