R S

• R S 6 May 2025 1:07 UTC

  3 points

  1

  on: Why Should I As­sume CCP AGI is Worse Than USG AGI?

  I think people focus too much on “would US AGI be safer than China” and not as much on “how much safer”

  In the sense that US has 15% pdoom and China has 22%, this notion that everyone needs to get onboard and help US win with full effort could be bad

  Could be used (and arguably is currently being used) to be even LESS safe, and empower an authoritarian mercantilist behemoth state, and possibly invade other countries for resources

  And in general massively increase and accelerate pdoom simply on the idea that our pdoom is lower than theirs

• R S 23 Mar 2025 5:43 UTC

  1 point

  0

  on: So­cially Grace­ful Degradation

  I agree with this although it makes me think about company culture

  There is huge emergent value to some of the.. let’s call them “softer” communication approaches

  It becomes possible to get out of random suboptimal Nash equilibriums almost immediately

  People can give more to each other, and better receive feedback

  But I think the only way to do this is by having the type of people who already think in those terms and prefer them

  There’s not a lot you can do to enforce it

  But it’s still a thing, and in my opinion it’s still a thing worth moving spaces for (to the extent that you have the luxury to do so)

  I don’t know that it scales very well, but at a certain scale it’s an incredible thing

  I’m currently leaving a job over this

  There are still some people who I know are capable of engaging in this way

  But we’ve kind of lost the battle, we lost too many critical people in our management chain

  So everyone is kind of reverting to these shittier communication styles

  It’s a bummer, and it just slows everything down so much

• R S 21 Mar 2025 18:28 UTC

  1 point

  0

  in reply to: kwiat.dev’s comment on: I changed my mind about orca intelligence

  I’m getting strong DarqqWolf vibes to this whole saga

  Although I’ve been that guy like 15 yrs ago, so I don’t fault him for it that much

  And at least he’s pushing a fairly novel idea and taking it seriously

• R S 19 Mar 2025 16:42 UTC

  1 point

  0

  on: Osaka

  Everyone is more likely to move to cities

  I vaguely remember that some physicist calculated that the density of everything you want increases by 15% every time a city doubles in size

  So there’s essentially a gravitational pull to cities

  Not unlike how (unconstrained by gravity) sea creatures continue to grow in mass because it’s more heat energy efficient due to the difference in scaling of volume to surface area

• R S 10 Mar 2025 23:54 UTC

  2 points

  0

  on: A Bear Case: My Pre­dic­tions Re­gard­ing AI Progress

  I agree with this insofar as this has always been my default /​ 60% case

  Selfishly I also hope this is how it plays out (for sake of my career)

  I also believe that it is the mainstream view

  But independently I think there’s a 20 to 30% chance that this is it, singularity hits very soon

  And I have to be prepared for that

• R S 11 Feb 2025 5:01 UTC

  1 point

  0

  in reply to: Raemon’s comment on: The “Think It Faster” Exercise

  Yeah I like that approach

  Part of it’s probably that I work very long hours often 7 days a week (blah blah stash money before ASI kills my differentiator)

  The biggest thing is not having a goal of being able to prove a solution to myself

  Like stop the search early basically

  Don’t need to formalize or document everything or remember every finding

  I think most of the time the brain actually does a pretty good job of gradually solving problems over time without conscious thought

  We don’t think of it as thought because it’s not conscious or subtitled by our auditory processing system

  But it still thought & learning

  Relying on that implicit system instead often gets me better results than logically formalized versions

  I guess now I’m kind of debating the core premise of the post—that it’s even possible or useful overall to discover or learn in that way, or that it’s real learning

  But it probably is helpful for some % of people who are less tunnely /​ tic-prone

  Maybe I’m just jealous that it works for some people after devoting (and largely wasting) so much time on it myself

  Not sure

• R S 9 Feb 2025 19:06 UTC

  7 points

  1

  on: The “Think It Faster” Exercise

  I basically discovered the same form of thinking after I learned the concept of amplification and distillation

  The long-term results of this was severe OCD that took me 1.5-2 years to cut back

  This is not to say that it’s a bad idea, or a bad idea for everyone

  But it’s a very, very bad idea for some people

  Basically I would spend so many hours every day just thinking, stuck in thought loops, stuck trying to gain some value out of generalizing

  Trying to compress mental models that are impossible to learn implicitly into simplified models/​memonics and then learn how to use them reflectively

  Etc etc

  It’s a fucking mess if you go too far

  And it’s hard to know if you’ve gone too far

  At least it was for me, until I was slowly working through it through months of therapy

  And then realized oh this is really all because of that stupid fucking thought pattern thing I’m obsessed with

  Especially the idea of trying to think it faster the FIRST time

  That in particular can be paralyzing, this sense that you should be able to think it the right way the first time

  Very slippery slope

  It’s like mindfulness, generally good up to a point, but it’s absolutely possible to go too far and have it cripple your life