so I've read a couple of screens, the entire conversation circles the main issue here with generative models (I won't get deeper to the AI revolution, singularity or something like that): the old world got old already.
the current systems, the legal one, copyrights everywhere, tons of human systems and processes which are based on human generated text and images (soon, music, video, probably, voices), as some kind of guarantee, just got deprecated simultaneously.
Soon most of them will be targeted by scammers of some kind, exploiting the soft spot of someone or some process requiring or presuming that some output in text or image format, shoud have to be created by a human being.
Kids faking their home assignments are just the beginning, in a year or two, IF the LLMs similar to chatGPT get somehow permanently open to the public, ala Google Search, you'll find every process presuming text or image output as only a human capable task, owned.
Owned like in hacking systems, the society will hack itself to subsume to newer LLM resultant ways of doing stuff, all the legacy societal systems and processes.
Of course, things like copyrights are the canary here, and the status quo has been succesful till now in containing the LLM tsunami of societal change (there you have google, facebook, the chinese giant LLMs, doing nothing publicly with their even more advanced LLMs), but as openAI has demonstrated, the ability to translate LLM outputs to money is incredible.
Maybe the FAANG don't need 10 billions bucks, but for sure is a big planet with many players are just looking how openAI using fairly dated AI tech, has gotten themselves 10 billion dollars of VC, just by publishing some chatbot with autoscaling infra behind.
So it is easy to predict that money will lead the adoption of future LLMs, despite what the FAANG or even state nations can regulate, whoever will end having the upper hand by adopting LLMs or newer AI tech available in the near future, most probably will end changing faster their own societies, giving themselves a too good to discard edge over the rest of the planet.
So things are changing, these LLMs, as simple as they are, could be just like the first submarine cables sending telegraph messages under the Atlantic for the first time. 30 years later and you can't recognize most of the societal processes running into the total system of the world.