LessWrong 2.0 Reader

✨ I just donated 71.12 USD (100 CAD 🇨🇦) ✨

I'd like to donate a more relevant amount but I'm finishing my undergrad and have no income stream... in fact, I'm looking to become a Mech Interp researcher (& later focus on agent foundations) but I'm not going to be able to do that if misaligned optimizers eat the world, so I support lightcone's direction as I understand it (policy that promotes AI not killing everyone).

If anyone knows of good ways to fund myself as a MI researcher, ideally focusing on this research direction I've been developing, please let me know : )

"three cultures of self-criticism" https://roamresearch.com/#/app/srcpublic/page/zzRZnCLd_

• non-self-critical culture (Barbarians):
  • baseline assumptions:
    • people generally think they are okay and good, and they are generally right.
    • self-criticism is rare.
    • if someone is being self-critical, guilty, ashamed, etc, that indicates an unusual problem.
  • implications:
    • intense self-criticism will be taken as evidence of something actually wrong with the person -- either they really did screw up quite badly, or they have poor judgment.
    • criticism is direct and overt.
      • if someone objects to what you've done, they'll tell you straight out, and expect that this will clear the air and lead to a resolution of the problem.
      • "negative" judgments are not necessarily intended, or expected, to be painful; the listener may very well disagree with the judgment or find it helpful feedback.
      • as a corollary, nobody assumes that an ambiguous comment or facial expression is a hint at criticism or disapproval. The default assumption is that people are fine with you, that you're fine, and if there's a problem it'll become obvious.
• pro-self-criticism culture (Puritans):
  • baseline assumptions:
    • people are generally deeply flawed; we are constantly screwing up, sinning, etc. this is the universal or near-universal human condition, not something limited to unusually bad people. but it really is genuinely Bad and Not Okay.
    • people tend to be complacent -- by default we engage in far too little self-criticism. We are screwing up without knowing it. We let ourselves off the hook, make excuses for ourselves, ignore warning signs. It takes active, continual effort to be vigilant against our own flaws.
  • implications:
    • intense self-criticism and guilt is normative. virtuous people will not think well of themselves. in fact, if someone does think well of themselves, that means they're lazy and have low standards.
      • corollary: an intensely self-critical or guilty person is not assumed to be an unusually bad person or to have a mental health problem; they are just doing what we're all supposed to do!
    • criticism can be harsh and intentionally painful, because the assumption is that it needs to be "strong enough" to overcome natural human complacency
    • it's also common to read criticism into subtle or ambiguous signs. the assumption is that there are always more problems than the obvious ones; it's never safe to presume things are fine.
• counter-self-criticism culture (Therapy Patients):
  • baseline assumptions:
    • people generally are too self-critical. most people are basically fine but torture themselves over minutiae.
    • complacency -- failing to self-criticize enough about genuine faults -- is literally monstrous. complacent people are rare, and pathological; we might call them sociopaths. you absolutely would not want to be one, and you're almost certainly not.
    • "healing" or "growth" means learning to quiet the overactive inner critic. this is very difficult; people need help with it.
    • everybody always needs validation and reassurance that they're ok, and the kindest thing you can do for anyone is give them permission not to worry or self-criticize. the cruelest thing you can do is trigger their insecurities and intensify their (already painful) self-criticism.
  • implications:
    • self-criticism is not normative; it's an affliction we all suffer from and long to be freed from.
      • like sin in pro-self-criticism cultures, misery in counter-self-criticism culture is seen as Genuinely Terrible, Deeply Not Okay, but also a part of the human condition, not a sign that something has gone unusually wrong with you. you're mentally ill, like everyone else.
    • criticism is mild and gentle, or suppressed altogether, because it's assumed everybody is already torturing themselves and doesn't need other people piling on.
      • corollary: it's common to read a lot of criticism or disapproval into subtle or ambiguous signals because it's assumed that people are holding back their true negative opinions. The absence of reassurance or validation is considered a sign of severe, harsh disapproval.
• relationships:
  • Barbarians see Puritans as totally excessive, and see Therapy Patients as trying to counteract a problem that one can just...not have.
  • Puritans see both Barbarians and Therapy Patients as dangerously complacent.
  • Therapy Patients see Puritans as a familiar enemy -- something they understand but reject and want to get away from, like an unhappy childhood home -- and see Barbarians as incomprehensible, alien, insane, not-even-human. 

The thing that comes to mind, when I think of "formidable master of rationality", is a highly experienced engineer trying to debug problems, especially high-urgency problems that the normal customer support teams haven't been able to handle.  You have a fresh phenomenon, which the creators of the existing product apparently didn't anticipate (or if they did, they didn't think it worth adding functionality to handle it), which casts doubt on existing diagnostic systems.  You have priors on which tools are likely to still work, priors on which underlying problems are likely to cause which symptoms; tests you can try, each of which has its own cost and range of likely outcomes, and some of which you might invent on the spot; all of these lead to updating your probability distribution over what the underlying problem might be.

Medical diagnostics, as illustrated by Dr. House, can be similar, although I suspect the frequency of "inventing new tests to diagnose a never-before-seen problem" is lower there.

Gaby, it seems, cannot imagine any reason one might think that children are good or that the country would be better off with more of them. They couldn’t mean what they say about demographic collapse and our dependence on growth. They couldn’t be genuine in their values. It must be a political takeover, or racism.

To understand either Gaby Del Valle's reaction or your reaction to Gaby Del Valle's reaction, it would be helpful if you wrote anything about the conference. Judging based on my prior, it is likely that people organizing such a conference are not orthogonal to other general political ideology. 

(My datapoint: people promoting more births etc were invited to different podcasts and it was mentioned they had a podcast of their own as well, so I searched that and was surprised that it seemed to be filled with almost cartoonish politics.)

On my understanding, the push for centralization came from a specific faction whose pitch was basically:

• here's the scaling laws for tokamaks
• here's how much money we'd need
• ... so let's make one real big tokamak rather than spending money on lots of little research devices.

... and that faction mostly won the competition for government funding for about half a century.

The current boom accepted that faction's story at face value, but then noticed that new materials allowed the same "scale up the tokamaks" strategy to be executed on a budget achievable with private funding, and therefore they could fund projects without having to fight the faction which won the battle for government funding.

The counterfactual which I think is probably correct is that there exist entirely different designs far superior to tokamaks, which don't require that much scale in the first place, but which were never discovered because the "scale up the tokamaks" faction basically won the competition for funding and stopped most research on alternative designs from happening.

I completed it.

I wish I read this sooner. Do you have a prototype or does this exist yet?

You're welcome!

Last year I had a version of that question where (mimicking a question the LW team asked) I said I'd keep it private. Reading the answers felt nice, and I realized an anonymous but public version of that could be really nice for a lot of people.

In general, at any given level of child maturity and parental risk tolerance, devices like this watch let children have more independence.

What has changed over the last few decades is primarily a large decrease in parental risk tolerance. I don't know what's driving this, but it's probably downstream from increasing wealth, lower child mortality, and the demographic transition.

I think the passphrase got introduced when there was a monetary reward attached at one point, and then reused for some prisoner's dilemma questions later.