Do you model Aurelia/Nectome as understanding this point?
kman
Karma: 1,655
Once the brain is dropped below freezing, I suspect that these vital areas are then shredded by ice.
How confident are you the shredding can’t be inverted?
This in isolation doesn’t seem like evidence about returns in the sense of what intelligence lets you accomplish that’s of strategic importance. It is evidence that humans vary a lot in cognitive ability as can be pretty directly measured by a simple puzzle, but not about the mapping of cognitive ability → strategic outcomes. You’d need to incorporate other bits of evidence for that.
“Constantly being sick as a baby/toddler or not” does seem like the sort of thing that could make a few IQ points worth of difference. Some things I found with Claude on a quick initial search:
This study found that a single infection resulting in hospitalization before the age of 19 was associated with an IQ 1.76 points lower at age 19, which shrunk to 1.13 points after adjusting for some confounders. They also found that the effects of these severe infections stacked, moreso if subsequent infections were of different types.
This one found that a single infection resulting in hospitalization before the age of 13 was associated with an IQ 1 point lower at age 18; 1.61 points lower if the infection was before the age of 1. The effects shrunk a lot after the age of 1: −0.7 points for ages 2-4, −0.26 points for ages 5-9, −0.12 points for ages 10-13. These effect sizes come from a model controlled for some confounders.
Another study found that having had any acute respiratory illness (ARI) before age 1 was associated with a 3 point lower Bayley score at age 1, but any ARI before age 2 was not significantly associated with score at age 2. (The Bayley is basically an IQ test for babies. If done at age 2 it probably correlates with adult IQ at something like r=0.3; even less if done at age 1.)
A study with a negative result: infections recorded between birth and age 5 were not significantly associated with IQ at age 11.
(Note: these could have been selected for positive results, either through publication bias or Claude searching.)
Sounds like you don’t actually disagree with “you build half as much tolerance” then?
My model says that if you took 400mg EOD:
on the caffeine days you’d feel similar to having taken 200mg in a caffeine-naive state
on the off days you’d feel similar to abstaining from caffeine after having taken 200mg daily for a while
200mg in a caffeine-naive state would feel pretty stimulating for most people I’d think. Do you feel more tired than a caffeine-naive baseline on the off days?
Sounds about right. I think it was a pretty dumb thought a day later.
I’ve been trying to compare the early/peak effects, which for caffeine will still be mostly just caffeine.
Also noticing I shifted the goalposts here from “stop giving a hoot” to “wouldn’t give as much of a hoot”. I concede the original point as worded.
Example that maps better: people probably care less about things said by someone that aged poorly, since there’s been a huge flood of such things due to social media.
Not sure I understand you. So for example the US public seems pretty desensitized to the US executive having constant scandals and being blatantly corrupt, and that’s basically one big important actor doing lots of clearly really bad stuff. If there’s a constant flood of private communications being leaked I’d guess people would get really desensitized to it (as well as any particular leak being drowned out by the rest of the flood). So it would get less useful as a weapon because the public wouldn’t give as much of a hoot, is what I was trying to say.
One thought that occurs to me is that, insofar as you’re going to be a target anyways, you should put yourself in the same class as the largest possible number of people, where you’re more likely to have recourse once that class is compromised. E.g. make sure you’re getting all the latest security updates on all your devices even if these are still vulnerable to zero days or supply chain attacks, so you don’t end up as one of the poor fools that got hacked for using some particular outdated thing.
You can maybe try to avoid having any important information with orgs/software that you don’t expect to be running the leading edge not-yet-public compsec AIs over their code.
I’m interested in thinking about how equilibria are going to shift. E.g. I think people will care a lot less about blackmail if it becomes ubiquitous.
and remember the West Anglia or Hillary or Epstein emails, how badly even the most innocent communication could be abused by fanatics or fools
There’s going to be so much of this over the coming years that I’m guessing people will be desensitized and stop giving a hoot.
I assume this is 1 reason a lot of people got addicted to nicotine: despite the shorter half life you need a longer time off to reset the effectiveness
I think nicotine is a strong behavioral reinforcer (especially when smoked, since it hits the brain immediately).
Caffeine: every other day likely keeps the same effect size?
My guess is that you build half as much tolerance taking X mg every other day compared to taking X mg daily.
Probably less of a kick if anything, it does seem a bit “smoother”. I also rather enjoy coffee, so not sure I’ll be replacing it long term.
Good point, I’ve been comparing to coffee. There could also be placebo effects since I associate coffee with stimulation. I’m also not sure how much I should trust the claimed caffeine content of the coffee pods I was using. I’ll update the post to mention these.
Where did you get the 100mg version?
here (The serving says 200mg, but that’s two capsules. They’re the only 100mg capsules I was able to find.)
Thanks so much for figuring this out and sharing.
You’re welcome! 🙂
I’m guessing this is assuming that caffeine and paraxanthine both have a half life of 5 hours, and that all caffeine is converted into paraxanthine. My simulator by default uses 4 hours for paraxanthine and 5 hours for caffeine, and a ratio of 0.84 for caffeine → paraxanthine.
My guess is this would be really underconstrained, maybe even if you had a ridiculous amount of data. The output you produce at some moment is a function of all past inputs. Brain state at that moment screens off past inputs, but then you’re trying to infer state at that moment with one piece of data, which is different from states at different moments.