I'll take the words of engineers and the people who work with these models over the words of some pseud who appeals to wannabe intellectuals with his use of philsophical buzzwords

The reason why you't really argue against his points is because he presents scenarios that haven't been proven. Kind of reminds me of the alarmism of how the research over at cern could end the universe. A lot of things can happen, I remember a couple years back there was a lot of talk about how AI was going to replace blue collar jobs first and now it's opposite.

The future is unpredictable and there's no point in trying to prevent scenarios that haven't happened.

If you want a good expert in the more non-technical side you should start with David Deutsch who actually has good credentials.

GPT-4 won't be that much bigger than GPT-3 according to Sam Altman, it'll still be bigger but not by that margin.

There is a neural network out there that has 500 billion parameters, but it's performance is still lower than neural networks with fewer parameters.

https://towardsdatascience.com/gpt-4-is-coming-soon-heres-what-we-know-about-it-64db058cfd45

The technology behind GPT-4 will be one of the many factors behind the big change that's coming. The advancements and milestones we are seeing in many fields of science and technology is what will change the world order by the next decade or two it won't just be Gptchat. I believe everything is connected with each other.