3 ms·

"Industrialism" (I lack a better term) is the real bane here. But it's sort of more than that, right? At the macroscopic level I think it's fair to analogize h

"Industrialism" (I lack a better term) is the real bane here. But it's sort of more than that, right? At the macroscopic level I think it's fair to analogize humanity as a system searching for equilibrium. But we can also do work on ourselves through technical and cultural revolutions. My foray into the past, watching the present, and my projections of the future all paint the classic "monkeypaw" trope. We cure childhood disease and population booms, increases the breadth of the economy, consumes masses of fuel and component elements while jointly polluting everything... We destabilize the system, creating novel problems of various sclae, and it just repeats ad nauseum. It doesn't matter what system you use, as long as we're complicating things as such none of them will ever "succeed" because they require some ideal state approximation that we will never let arise. In engineering everything is a compromise, you can't have your cake and eat it too, etc... Neither Marx nor Friedman are really wrong, per se, but they can't be right, either. From there it's just a matter of moral convictions, at least in my opinion, which paints everything with an even less discernible gray.

2moreweeks 3y ago
Quoting: "We cure childhood disease and population booms, increases the breadth of the economy, consumes masses of fuel and component elements while jointly polluting everything... We destabilize the system, creating novel problems of various sclae, and it just repeats ad nauseum. It doesn't matter what system you use, as long as we're complicating things as such none of them will ever "succeed" because they require some ideal state" Yesterday, i saw people disputing at the internet. I'd not read nor join their conflict, but made my jokes about. For today i intended to read, how and what they said. Quoting (another Topic on HN) some Posts more: "Further: progress will occur on real-world problems that can be solved by those methods, making those applications dominant throughout society." (especialy everything looks like its natural to someone) Jesssus! They need a pulling-horse, that dies... (-; "Taste was bitter at the time, but it was never going to be forever. In 2018 models had 100e6 parameters, and in 2013 GPT4 has (wild guess) 2e12 parameters [0]. Conservatively that's 500% per year. At it's peak, hardware never improved at that rate and has since slowed down dramatically. So it was always going to end, and further advancement was always going to revert being driven by a neural net of some sort." The interesting question when neural net will end up driving it. "Posits that making a huge model won't be a competitive strategy for long because it gets to expensive in terms of heavy quantization." # ...you may also like to read about: "KYBERNETIK", "Prozessablaufmaschinen", Bottleneck... (calms) regards...

2moreweeks 3y ago

Quoting: "We cure childhood disease and population booms, increases the breadth of the economy, consumes masses of fuel and component elements while jointly polluting everything... We destabilize the system, creating novel problems of various sclae, and it just repeats ad nauseum. It doesn't matter what system you use, as long as we're complicating things as such none of them will ever "succeed" because they require some ideal state" Yesterday, i saw people disputing at the internet. I'd not read nor join their conflict, but made my jokes about. For today i intended to read, how and what they said. Quoting (another Topic on HN) some Posts more: "Further: progress will occur on real-world problems that can be solved by those methods, making those applications dominant throughout society." (especialy everything looks like its natural to someone) Jesssus! They need a pulling-horse, that dies... (-; "Taste was bitter at the time, but it was never going to be forever. In 2018 models had 100e6 parameters, and in 2013 GPT4 has (wild guess) 2e12 parameters [0]. Conservatively that's 500% per year. At it's peak, hardware never improved at that rate and has since slowed down dramatically. So it was always going to end, and further advancement was always going to revert being driven by a neural net of some sort." The interesting question when neural net will end up driving it. "Posits that making a huge model won't be a competitive strategy for long because it gets to expensive in terms of heavy quantization." # ...you may also like to read about: "KYBERNETIK", "Prozessablaufmaschinen", Bottleneck... (calms) regards...