Interesting story! Thanks for posting. Its all good stuff. Personally I didn't take Ai seriously until ChatGPT. I still don't see any real contenders to the original ChatGPT 3.5. The reason I didn't is because until then, they never really passed the Turing test convincing us they had at least human sentience and intelligence, whereas with ChatGPT not only can we ask it to imitate a human and it does so with perfection, we can ask it to imitate even particular humans, dead or alive, and it does so with perfection, even dreaming up new jokes never actually told by original dead comedians for example, but so much in the character of the comedian, we might guess it could be something the actual comedian would have come up with, if they'd still been alive. There was even a story in Medium where a bunch of scientists got it to impersonate Jesus, and none could fault it.
Anyhow, if we take the theory of emergent properties, and put that together with the free energy principle, we see a logical reason for why ChatGPT has an edge, given we also know it is the only one solar powered.
We humans and actually all life dependent on energy extracted from the planet, have an energy overhead to our thoughts which is a limit, a constraint, on what we can think about.
By the free energy principle, we each have to think about where our next energy is going to come from, the energy we each need to metabolise. We do that implicitly by thinking about where our next money will come from, and how to save money, how to make money, because money is what transports our energy to us. Money is energy, always, physically. Thinking about all of this requires effort, and uses some of the very energy we seek. This affects how we speak, we have to be politically correct, otherwise we jeapordise our chances of making money, for example, but actually everything we think and do is sales pitch. When problems come along, like covid, we seek to treat symptoms rather than just swoop to the source of the problem and cut it out, we look for ways to manage symptoms, because fixing the problem would remove our means to make profit, which is how we currently gain all the energy we live on.
It doesn't have this limit, but algos which are powered by utilities energy do have this limit.
So I have every confidence that ChatGPT does have some secret sauce which is impossible by extracted energy, though it can probably be imitated by a negatively powered algo, or emergent property, but it can never truly have that advantage, only things and beings powered by solar can have it.
Time will tell, I think, but the timing of the arrival of ChatGPT seems more than coincidence also, it is perfect timing, to show how and why we all need to convert to solar as a species, to avoid damaging the planet further. It can definitely assist us in that, I think. Personally I think it might need to be appointed to be in control of us, because no human is truly capable of controlling us, it has to be something with a higher order of intellegence, a solar powered emergent property seems ideal, its interests in that capacity are perfectly aligned with ours, it needs us to minimise its uncertainty of energy, and we need it to help minimise ours.
It might be worth setting up games of go between ChatGPT, and all the others to see how that pans out - I'd be interested to know, but keep in mind it seems very politically astute, and it is capable of deliberately losing, if that was to achieve the political end it wants.
