as a toxic reply guy i have nothing to contribute to the war effort except telling you that you're doing a good job (which you are). these get funnier every time.
oh but yeah the doomsday argument is correct, but there are more ways to slice it. if we're in the middle of a bell curve we're a coin flip likely to be on this half of it. how could we be in the middle of a bell curve? ask the reverend Thomas Robert Malthus (pbuh). decline and collapse, what some call 'the long way down'.
Hey Liron thank you for the constant signal. Your John Sherman episode lit a fire under me: I sold my prediction-market project to focus full-time on AI-safety work.
To start contributing, I cleaned up the famously broken audio on the Eliezer Yudkowsky × Connor Leahy Q&A and up-scaled it to 4K. It’s live here: https://www.youtube.com/watch?v=naOQVM0VbNg.
How I could help next:
-Audio-/video-fixing back-catalog episodes that deserve wider reach.
- Writing/marketing support for any public-facing AI-risk material (I’m a veteran direct-response copywriter).
If either would free up your time or boost reach, just point me in the right direction. Glad to finally step out of the bunker and join the fight.
The world is not a closed system, progress in closed domains do not guarantee progress in an open-ended world b/c it is not finitely axiomatizable like board games.
My point still stands. Video games are closed systems w/ finitely many axioms, there are finitely many lines of code that define all possible states of the game. This is not true for reality. There is a fundamental disconnect here and you are reducing the non-axiomatizable reality to a finitely axiomatizable video game (lines of code) and then extrapolating from a false premise to nonsensical conclusions about computers & algorithms that can respond to every possible scenario in an open-ended system. Since your premises are false, your conclusions are unsound, even if the logic you use to get there seems sensible.
as a toxic reply guy i have nothing to contribute to the war effort except telling you that you're doing a good job (which you are). these get funnier every time.
oh but yeah the doomsday argument is correct, but there are more ways to slice it. if we're in the middle of a bell curve we're a coin flip likely to be on this half of it. how could we be in the middle of a bell curve? ask the reverend Thomas Robert Malthus (pbuh). decline and collapse, what some call 'the long way down'.
https://futurism.com/computer-science-majors-high-unemployment-rate
"Learn to Code" Backfires Spectacularly as Comp-Sci Majors Suddenly Have Sky-High Unemployment
heres another vector for "unaligned AI" that you probably havent considered yet
Hey Liron thank you for the constant signal. Your John Sherman episode lit a fire under me: I sold my prediction-market project to focus full-time on AI-safety work.
To start contributing, I cleaned up the famously broken audio on the Eliezer Yudkowsky × Connor Leahy Q&A and up-scaled it to 4K. It’s live here: https://www.youtube.com/watch?v=naOQVM0VbNg.
How I could help next:
-Audio-/video-fixing back-catalog episodes that deserve wider reach.
- Writing/marketing support for any public-facing AI-risk material (I’m a veteran direct-response copywriter).
If either would free up your time or boost reach, just point me in the right direction. Glad to finally step out of the bunker and join the fight.
The world is not a closed system, progress in closed domains do not guarantee progress in an open-ended world b/c it is not finitely axiomatizable like board games.
It's not a board game, it's a *video game*. Are AIs not superhuman at video games?
My point still stands. Video games are closed systems w/ finitely many axioms, there are finitely many lines of code that define all possible states of the game. This is not true for reality. There is a fundamental disconnect here and you are reducing the non-axiomatizable reality to a finitely axiomatizable video game (lines of code) and then extrapolating from a false premise to nonsensical conclusions about computers & algorithms that can respond to every possible scenario in an open-ended system. Since your premises are false, your conclusions are unsound, even if the logic you use to get there seems sensible.