jmh

• jmh 27 May 2025 15:48 UTC

  2 points

  0

  in reply to: David Corbitt’s comment on: Ori­ent­ing Toward Wizard Power

  I’m not sure I’m understanding your claim here. I don’t think you’re suggesting a “if you build it they will come” type result from being in a dominant position. But I also don’t quite get why being in a dominant position in one space has much (other that deep pockets) really impacts the success in doing things others are not doing—or leading a group in doing things others are not leading that group.

• jmh 25 May 2025 0:09 UTC

  4 points

  0

  in reply to: DirectedEvolution’s comment on: That’s Not How Epi­ge­netic Mod­ifi­ca­tions Work

  Probably not the type of reference you were thinking about regarding reprogramming and impact on aging issues but I suspect it’s in the area you were thinking. I’m pretty sure it’s been mentioned here on LW before in other posts/​comments. Interesting idea but implementation is problematic to say the least—but really hoping someone can figure it out.

  Since a lot of this is way beyond my skill sets and knowledge, when you’re looking at the dynamic interplay aspect, is that purely internal to the cell or do you also look at the extra-cellular “communications”? If so, are you familiar with the Conboy’s plasma dilution experiments?

• jmh 17 May 2025 1:51 UTC

  3 points

  1

  in reply to: Chipmonk’s comment on: So­cial Anx­iety Isn’t About Be­ing Liked

  Yes, I agree. But actions that make people dislike you less may have no impact on people liking you or may have a positive impact (increase you likeability). How does one go about choosing which element from the action set to put in your model to produce the results you present?

  I suppose a better way to put my take here is that you’re presenting a limited/​narrow model but the rhetoric implies some type of general model. I think that is a bit dangerous to people trying to both understand their social anxiety and do something if they are looking to change it (which seems a bit implied by the use of anxiety rather than just saying introverted).

• jmh 17 May 2025 1:06 UTC

  4 points

  −2

  on: So­cial Anx­iety Isn’t About Be­ing Liked

  I like the general thrust of the post here but not sure I agree with some of the assumptions and conclusions that seem to follow from assumptions. I think you’re correct that there as some real asymmetries that get lost when confusing avoiding dis-like with getting more liked. But I think that comes from disliked and liked not being different ends of a common spectrum. I suspect they are somewhat orthogonal to one another.

  So the distribution you draw is imposed on a projection of the dislike vector and the like vector which can allow shifting from the insecure to the secure distribution producing no change in the like number in the tail, or arguably no change in the dislike number, with an increase in the like.

• jmh 14 May 2025 14:10 UTC

  2 points

  0

  on: The Best Refer­ence Works for Every Subject

  I would add best literature survey piece for the given subject area as another type of reference material people should submit.

• jmh 14 May 2025 14:02 UTC

  5 points

  0

  on: Too Soon

  My sympathies to you and your family. This is one of life’s sadness that we all experience at some point. While I don’t think this will help easy anything (what could I say to do so?) I have to say when may father died it was too soon—or felt that way. He lived a long life, died from aging (at nearly 102) and lasted a few days after the pain got too much for him that morphine was regularly administered by the hospice nurse. It was time, and expected. When it happened it was still too soon.

  It was easier for me when my mother died about 6 months later but I didn’t not have the same sense of loss, being lost, wishing there was more I had done, regrets regarding the relationship and more. And in her case there was even a bit of your experience of an initial mis-assessment (in our case be a sybling) that was the final straw with my mother dying a couple of weeks later.

  I hope your father is taking it well (that sounds bad because I know it’s nothing but pain, but I cannot think of any better way to say) and has other reasons to view life as worth living rather than feeling that same deep too soon loss and a lack of interest in what life offers.

• jmh 14 May 2025 0:23 UTC

  2 points

  0

  in reply to: nim’s comment on: Ori­ent­ing Toward Wizard Power

  Perhaps a more general way to approach the question would be can one identify the existing comparative advantages related to the wizard power related task to be performed.

  If you only want parts which are easy to clearly describe, you can hire someone to CAD them up… or you could describe them to an LLM and have it emit files that you can convert to your CAD of choice.

  Don’t know if this is still the case but 15 years ago something very similar existed with CAD/​CAM production. The CAD design could be passed through a processor to generate the G-code (instructions the CAM processor reads and follows) and then some machinist would review and “fix” tool path or order to make the code more efficient for production.

• jmh 12 May 2025 9:49 UTC

  2 points

  0

  on: a con­fu­sion about prefer­ence orderings

  The inference goes beyond the definition and claims, additionally, that in a case like this diagram, the agent will only take actions that “follow the arrows,” changing the world state from less-preferred to more-preferred at each step.

  Hmm. I’m not sure about that inference but might be due to absence of specific argument you’re making that inference from.

  The preference arrow to me is simply pointing in a direction. The path between the less preferred state to more preferred state is pretty much a black box at this point. So I think the inference is more like: If agent 86 stays in state A, there is a lack of resources or a higher transition cost to get from A to C than a difference in utility/​value between the two states.

  There is probably something of a granularity here as well. Does taking a second job for a while, or doing some type of gig-work one doesn’t enjoy at all, count as a state or is that wrapped up in some type of non-preference type of state?

• jmh 11 May 2025 7:41 UTC

  3 points

  0

  in reply to: Dweomite’s comment on: Ori­ent­ing Toward Wizard Power

  Perhaps it is more terminology as I didn’t mean to suggest some existing tendency was merely coincidental. I suspect, and would agree with the points you make, that some underlying structures will produce that fewer king opportunities and more wizard opportunities world.

  I see a lot of complications in the interactions here related to the kings and wizards framing that can muddy things up. But was really just observing that these are not quiet as simple as choose wizard power over king power. I think one finds more wizard opportunities (to be found within all parades, between parades and even before/​after/​separate from parades), but think that sometimes we’ll have new parades that want to do their thing but existing kings are able to see that path so a new king (not really competing with other kings as the parade will not follow them far) will add a lot of value. That then opens up more wizard opportunities. Plus, I suspect many wizards would be poor kings and many kings poor wizards—so think about who you are when looking for you “power”.

  Choosing the wrong option or choosing one at the wrong time will be low/​negative value (probably both personally and systemically). But if you’re unsure about it, choosing wizard is probably the smarter choice.

• jmh 10 May 2025 1:53 UTC

  3 points

  0

  in reply to: David Lorell’s comment on: Ori­ent­ing Toward Wizard Power

  Do you have an indications that those without the clinical signs of depression (or at least doctor approved state) won’t become acclimated to the drug in a way those that perhaps need it for a balanced state don’t?

  I suppose asking a bit differently here, what are the gears here that do the work and how well one might think they they match up with one’s own system that is in place?

  Interesting though about using it to improve one’s performance rather than just as an antidepressant or aid to quit smoking. the wiki has some good info but interesting that it doesn’t have a strong effect on dopamine so makes me wonder if looking more at what norepinephrine does, or perhaps the ratio between norepinephrine and dopamine.

  Any consideration on the use of other NDRI drugs rather than Bupropion. I’ve not looked into much at this point but Bupropion does have some side effects I would not be too interested in experiencing.

• jmh 9 May 2025 16:34 UTC

  9 points

  0

  in reply to: Dweomite’s comment on: Ori­ent­ing Toward Wizard Power

  Interesting distinction to draw out. I some what come away with a view that wizards get to play in the positive-sum game world while kings are stuck in a zero- to negative-sum game. I’m not completely sure that is a good modeling though might see the game of thrones hitting it’s equilibrium at a lower number of kings than one might expect to occur with wizards (resources are not infinite so at some point wizards will also just be competing against one another for stuff—and possible followers for their own parades).

  So maybe the bit is that we will tend to find ourselves in a world where kings are at, or past, the equilibrium level, but the market for wizards will still support expansion and growth. I suspect there is likely some interaction here such that as wizards “create” and more wizards then create more, this activity produces some scope for increasing the kingdoms.

  So perhaps it’s not really about king power or wizard power but identifying where the greatest lack resided.

  I’m not sure that is the whole story here though as some of John’s post seemed to point at purely personal aspect—where that lost part went and why. I’m not sure that the king/​wizard framing gets to the core of that.

• jmh 7 May 2025 0:19 UTC

  5 points

  0

  on: Five Hinge‑Ques­tions That De­cide Whether AGI Is Five Years Away or Twenty

  Current AI, even after ChatGPT, earns roughly ten‑thousand dollars per H100‑GPU‑year.

  What does that $10K number actually represent? An average across all AI? The marginal GPU earning across all AI? An estimate of either from one of the big AIs?

• jmh 27 Apr 2025 2:11 UTC

  3 points

  1

  in reply to: Alexander Howell’s comment on: Why Should I As­sume CCP AGI is Worse Than USG AGI?

  I always find the use of “X% of the vote” in US elections to make some general point about overall acceptability or representation of the general public problematic. I agree it’s a true statement but leaves out the important aspect of turn out for the vote.

  I have to wonder if, particularly the USA, would not be quite as divided if all the reporting provided percentage of vote of the eligible voting population rather than the percentage of votes cast. I think there is a big problem with just ignoring the non-vote information the is present (or expecting anyone to look it up and make the adjustments on their own).

  But I agree, I’m not too sure just where electoral systems fall into this question of AGI/​ASI first emerging under ether the USA or CCP.

• jmh 27 Apr 2025 1:48 UTC

  2 points

  0

  on: To Un­der­stand His­tory, Keep Former Pop­u­la­tion Distri­bu­tions In Mind

  Would I be incorrect in reading into the post that, at least looked at through the lens of geopolitical status and power, one might not view the post WWII international order than promoted a lot of investment in the developing countries economic and health was planting the seeds of the West’s vision of the modern world order?

  Or do you see the current state of things as occurring independently of relative population changes?

• jmh 22 Apr 2025 13:06 UTC

  3 points

  1

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

  I mostly put this question through the same filter I do the question of Chinese vs. US hegemony/​empire. China has a long history of empire and knows how to do it well. The political bureaucracy in China is well developed for preserving both itself and the empire (even within changes at the top/​changes of dynasty). Culturally and socially the population seems to be well acclimated to being ruled rather than seeing government as the servant of the people (which I not quite the same as saying they are resigned to abusive totalitarianism, the empire has to be providing something that amounts to public good and peace but seem more tolerant of the means applied than western cultures).

  The US on the other hand pretty much sucks at empire and lacks a well functioning and well developed bureaucracy for supporting empire.

  So I think perhaps one might need to ask what the specific risk one is most concerned about. China probably produces an AI that serves the Party and Empire and so perhaps is a bit more corrigible and won’t decide to kill everyone. But if you’re concerned about what those in control of an AGI/​ASI might do with it China might be a bigger risk than the US. With the US you probably have greater risks from the AI wanting to kill everyone or simply doing that without caring but likely have more controls on what the AI will allow its “owners” to do with it.

• jmh 13 Apr 2025 1:50 UTC

  2 points

  0

  in reply to: Sergii’s comment on: Can I learn lan­guage faster? Or, per­haps, can I mem­o­rize the for­eign words and re­call them faster?

  That is pretty much what I’m trying to accomplish and want to try to increase the rate I am building the working vocabulary.

  I do agree with both you and Vaughn. Reading should (very hard for me now) really help improving the recall once I can read and have a sufficient understanding of the statement and larger text. Texting is (I have been able to do some) good for me in that it tends to keep the exchange short and sentence structure more simple and short (which means I typically will have a reasonable grasp of the general meaning so can better infer what the unknown word or unrecalled word likely means.)

Can I learn lan­guage faster? Or, per­haps, can I mem­o­rize the for­eign words and re­call them faster?

• jmh 4 Apr 2025 14:01 UTC

  6 points

  4

  on: Why Have Sen­tence Lengths De­creased?

  Literacy seems to make sense to me but I might be missing something in the post. Writing is language and language is communication so at least two sides.

  As more people learned to read, they also learned to write, and written communications increases. However, even with modest literacy one can read a long sentence. Or can do that when it is written by a good/​skilled writer. But being able to read does not really lead to writing skills in most cases I suspect.

  As more people started communicating via writing (think things like schools and education expansion) the skill level of the average writer likely declined. That probably lead to training next generation writes to write in a more simple sentence structure.

• jmh 28 Feb 2025 6:25 UTC

  5 points

  5

  in reply to: interstice’s comment on: You can just wear a suit

  There’s a Korean expression that basicly seems to be “the look is right” or “the look fits” which seems in line with your comment. The same outfit, hat, shoes, glasses, jacket or even car for different people create a different image in other’s heads. There is a different message getting sent.

  So if the overall point for the post is about the signaling then I suspect it is very important to consider the device one chooses to send messages like this. In other words, yes breaking some social/​cultural standards to make certain points is fine but thought needs to be put into just how appropriately your chosen device/​method “fits” you will probably have a fairly large impact on your success.

  I suspect that holds just as well if you’re looking at some type of “polarizing” action as a mechanism for breaking the ice and providing some filtering for making new acquaintances and future good friends.

• jmh 23 Feb 2025 20:02 UTC

  3 points

  −4

  on: How to Make Superbabies

  I’m reminded of the old Star Trek episode with the super humans that were found in cryosleep that then took over the Enterprise.

  While I do agree that this could be one potential counter to AI (unless the relative speed things overwhelm) but also see a similar type of risk from the engineered humans. In that view, the program needs to be something that is widely implemented (which would also make it potentially a x-risk case itself) or we could easily find ourselves having created a ruler class that views ordinary humans as subhuman on not deserving of full rights. Not sure how that gets done though—from a purely practical and politically viable approach.

  I certainly think if we’re doing things piecemeal we would want somewhat smarter people before we have much longer living people.