Dumbledore's Army

Karma: 904

Dumbledore's Army 17 Mar 2026 14:29 UTC
13 points
7
in reply to: cdt’s comment on: My Willing Complicity In “Human Rights Abuse”
I disagree that this is a fully general argument and I think the phrase “a quick welfare race-to-the-bottom whereby people can be subjected to anything as long as there is something to gain” is tendentious and hostile.

I do generally support letting people take ‘bad’ options which they reflectively endorse as their least-bad choice in the imperfect world we actually live in; AlphaAndOmega has gone into depth on the specific ‘bad’ option of performing migrant labour in Qatar. I can’t speak for AlphaAndOmega, but I’m also a libertarian and have strong moral objections to stopping consenting adults making their own choices which don’t affect anyone else.

But here are some things people might be subjected to which I don’t support, and which this “fully general” argument doesn’t support:
- Anything nonconsensual
- Deceiving people into giving consent
- Things which have large negative externalities on non-consenting third parties
- Getting minors or the mentally handicapped to ‘consent’ to things they are not equipped to understand
I suspect that you are pattern-matching us against some other argument that is not quite the one we’re making. Perhaps I’ve misunderstood you—would you like to elaborate on your concerns?

Dumbledore's Army 15 Mar 2026 15:09 UTC
28 points
10
on: My Willing Complicity In “Human Rights Abuse”
Thank you for an interesting and important post. It is amazing how many well-meaning Westerners want to ban poor and desperate people from doing things that would genuinely make their lives better, simply because those Westerners cannot imagine themselves ever being so poor and desperate that they would voluntarily [work as a labourer in Qatar | work in a sweatshop | sell a kidney | insert example here].

I made a similar point in Don’t take bad options away from people but I hope that your more concrete and less abstract post will persuade people more effectively.

Dumbledore's Army 25 Feb 2026 16:16 UTC
10 points
2
on: What was the most effective team you’ve ever been on, and what made it excellent?
Really good question, thank you. I look forward to hearing more answers.
For me, it was working as an investment analyst—sorry, I’m going to redact the company name to protect my identity. I’ve held roles at a few firms, but one stood out:
- Great clarity on the mission, which was investment outperformance.
- Extensive efforts to measure and track investment performance. (I want to say we started with an advantage, because investment performance is quantifiable in a way that lots of desirable outputs are not, but even within the investment industry, the level of focus on goal is very variable. One company I worked for was so bad that my team literally never had a discussion about our investment performance even though it should be the number one mission-critical item.)
- Thinking systematically about incentive structures. There was an investment team (including me), incentivised to focus on investment performance. We were not incentivised to focus on general company profitability, client satisfaction metrics, or other genuinely important stuff. We had a job, we were incentivised to do that job. There were other people incentivised to focus on various other things and make sure other jobs got done. (Importantly relevant: these jobs were separable. My team wasn’t picking up the phone to clients, so we didn’t need to have incentives about that. The actual client-relations team weren’t picking the investment portfolio. Etc.)
- Sensible design of the work environment. It was an open-plan office, but we had our own floor away from distractions like people yelling into phones all day.
- Limited number of meetings, clear purpose for each meeting we had.
- Hiring by the team doing the job. Ten years on, I literally can’t remember if we had a separate HR function. I think we just had an admin team for payroll and stuff. But the somewhat limited hiring (maybe 1 person per year on an 8-ish person team) was done by the team, who knew what technical competence looked like and who could detect bullshit.
A lot of that would be difficult to replicate by individuals, since it was heavily dependent on the company’s CEO and senior team being ultra-focused on actually achieving outperformance and not symbolic busywork, but potentially relevant to start-up founders.

Dumbledore's Army 20 Feb 2026 10:07 UTC
10 points
0
on: You May Already Be Canadian
I’m in a somewhat similar position as a British citizen with an Irish grandparent—I’m entitled to dual citizenship. I got the passport even though I live in the UK—the option value of being able to easily move to the EU is valuable. Given recent developments in the USA, I would think that the option value of having a plan B is even higher.

Relatedly, anyone with a single grandparent who was an Irish citizen born on Irish soil can get Irish citizenship, so there’s probably a good number of Americans who qualify. The application process was cheap and relatively easy. At the time it was slow—there was a long post-Brexit backlog—but that may have eased now.

Dumbledore's Army 6 Nov 2025 11:58 UTC
5 points
2
on: Heroic Responsibility
As several commenters here have said, the business owner example isn’t a great fit for heroic responsibility. The core is taking responsibility for things that aren’t your job, that you are not socially expected to be responsible for, because you have decided that the thing needs to be done.

The archetypal fictional example is the hero who raises the rebellion that overthrows the Evil Empire. A normal sensible peasant whose home has just been burned doesn’t do that, he just tries to survive the winter. The hero decides to do more than that, even though it’s not their job and they’re not (at the start of the book/movie/whatever) distinguished from all the other thousands of people in the same position. That choice is why we call them the hero.

Or, for a real-life example, a factory-owner in 1940s Germany might reasonably feel that he has quite enough to do trying to keep his business running in a war zone. It’s not his job to save Jewish lives, they’re not his relatives or anything. And even trying to save them could be very hazardous. Nevertheless, Oskar Schindler decided he was going save as many lives as he could anyway, and saved about 1,200 people. That’s what heroic responsibility looks like.

And, the most obvious example to anyone on LessWrong: there is no social expectation that a normal citizen who is not a senior government official will decide that it’s up to him to save the world from the risk that Unfriendly AI kills us all, but Eliezer Yudkowsky saw that no one else was going to do the job, and he wants the world to be saved, so he’s spent his career trying.

The heroic responsibility described by OP, the determination to focus on whether [goal] is actually happening or whether people are buck-passing or symbolically pretending to do the thing, or whatever other failure mode—all that is downstream of the heroic decision to take responsibility for [goal] in the first place.

Dumbledore's Army 15 Sep 2025 8:59 UTC
9 points
4
on: Does My Appearance Primarily Matter for a Romantic Partner?
I agree with others here that appearance is regrettably important in general.

You ask about looking good for a romantic partner. I suggest you should additionally think about screening effects. Remember that, as a generalisation, the typical woman faces a different problem than the typical man when looking for a partner: the woman has to screen through a large number of offers, many/most of which are low-quality, to find a partner she wants. People in general respond to screening problems with shortcuts and heuristics. A person who knows you well will consider your appearance as one aspect of what they know about you; a person who has just met you is likely to make a snap judgement on your appearance. Your aim should be for an appearance good enough that it is not making most potential partners screen you out at the first step.

Dumbledore's Army 29 Aug 2025 7:20 UTC
4 points
0
in reply to: bfinn’s comment on: The Worst Form Of Government (Except For Everything Else We’ve Tried)
I can’t think of such a term, and AI’s best effort is ultracrepidarianism, which isn’t quite the same, and which doesn’t roll off the tongue. Unless someone else has a better idea, I suggest ‘misgeneralising’. You’re right that it’s a relatively common bias and needs a name.

Dumbledore's Army 18 Apr 2025 8:36 UTC
2 points
0
on: The Russell Conjugation Illuminator
Thank you for creating this.

Dumbledore's Army 8 Jan 2025 21:14 UTC
1 point
0
in reply to: espoire’s comment on: The Online Sports Gambling Experiment Has Failed
Thanks for the comment, and I agree that it would be helpful if the debate on this topic could focus on the question of ‘how do we encourage people to take up a less harmful addiction’—like vaping vs smoking—rather than typically jumping to the question of ban / don’t ban.

Dumbledore's Army 8 Jan 2025 21:11 UTC
4 points
0
in reply to: alexey’s comment on: The Online Sports Gambling Experiment Has Failed
“Except whatever they got addicted to before the legalization of online sports betting, it apparently led to much lower bankruptcy rates etc.”
Yes, Zvi gives evidence on bankruptcy rates. However, that is not the only kind of harm. Sports gambling doesn’t have direct health effects the way that drug or alcohol addictions do. Different types of addictions hit relationships, and sports gambling isn’t good for relationships, but it’s plausibly less harmful than a porn addiction. Sports gambling leaves people functional and able to hold down a job, again in a way that not all other addictions do. I don’t think you can assert that banning sports gambling will make people ‘much less harmed’. Not unless you’ve done a deep dive into all the different forms of harm caused by different types of addiction which they might get instead. If someone like Scott Alexander did that kind of analysis and announced that banning sports gambling is still worthwhile because people will addict themselves to cannabis or porn and that is net better, that would be different, but so far as I know, no one has tried to answer the question.
Separate point: you also can’t make the jump from ‘sports gambling is harmful’ to ‘we should ban sports gambling’. Banning things also causes harm. It moderately reduces takeup, but doesn’t eliminate the thing banned—illegal gambling is a problem everywhere people ban it, Prohibition didn’t eliminate alcohol, and the War on Drugs didn’t magically stop people taking them. So banning things causes a moderate reduction in usage of the thing, with an empirical question mark over how much. Banning things also means criminalising people who would otherwise not be criminalised, ruining their lives. It means devoting more societal resources to police and prisons, to enforce the ban, or else it means diverting existing law enforcement resources away from dealing with other crimes. Even if you accept both that sports gambling is harmful to users (more harmful than whatever other thing they’d get addicted to instead) and that banning it would somewhat reduce usage of gambling, you still can’t come to a firm conclusion that banning it is net beneficial to society.
On Zvi’s levels of legality, I’m OK with the idea of making sports gambling a 2, ie legal with modest frictions, but I really don’t believe that society can justify going to level 3 and criminalising it. And, again, the whole should-we-ban-it debate is a distraction that doesn’t help with the broader question of how society can treat people with addictive personality disorder.

Dumbledore's Army 19 Dec 2024 17:02 UTC
7 points
1
in reply to: Julian Bradshaw’s comment on: Dark Forest Theories
Hard disagree to point 1. The fact that humanity hasn’t tried to hide is not counter-evidence to the Dark Forest theory. If the Dark Forest is correct, the prediction is that all non-hiding civilisations will be destroyed. We don’t see anyone else out there, not because every civilisation decided to hide, but because only hiders survived.

To be clear: the prediction of the Dark Forest theory is that if humanity keeps being louder and noisier, we will at some point come to the attention of an elder civilisation and be destroyed. I don’t know what probability to put on this theory being correct. I doubt it ranks higher than AI in terms of existential risk.

I do know that ‘we haven’t been destroyed yet, barely 100 years after inventing radio’ is only evidence that there are no ultra-hostile civilisations within 50 light-years which also have the capability to detect even the very weakest radio signals from an antique Marconi. It is not evidence that we won’t be destroyed in future when signals reach more distant civs and/or we make more noticeable signals.

Dumbledore's Army 14 Dec 2024 10:51 UTC
9 points
−20
on: The Online Sports Gambling Experiment Has Failed
A parable to elucidate my disagreement with parts of Zvi’s conclusion:
Jenny is a teenager in a boarding school. She starts cutting herself using razors. The school principal bans razors. Now all the other kids can’t shave and have to grow beards (if male) and have hairy armpits. Jenny switches to self-harming with scissors. The school principal bans scissors. Now every time the students receive a package they have to tear it open with their bare hands, and anyone physically weak or disabled has to go begging for someone to help them. Jenny smashes a mirror into glass shards, and self-harms with those. The principal bans mirrors...
Any sane adult here would be screaming at the principal. “No! Stop banning stuff! Get Jenny some psychological treatment!”
Yes, I know the parable is not an exact match to sports betting, but it feels like a similar dynamic. There are some people with addictive personality disorder and they will harm themselves by using a product that the majority of the population can enjoy responsibly^[1]. (Per a comment above, 39% of the US population bet online, of whom only a tiny fraction will have a gambling problem.) The product might be gambling, alcohol, online porn, cannabis, or something else. New such products may be invented in future. But, crucially, if one product is not available, then these people will very likely form an addiction to something else^[2]. That is what ‘addictive personality disorder’ means.
Sometimes the authorities want to ban the product in order to protect addicts. But no one ever asks the question: how can we stop them from wanting to self-harm in the first place? Because just banning the current popular method of self-harming is not a solution if people will go on to addict themselves to something else instead^[3]. I feel that the discourse has quietly assumed a fabricated option: if these people can’t gamble then they will be happy unharmed non-addicts.
I am a libertarian, and I have great sympathy for Maxwell Tabarrok’s arguments. But in this case, I think the whole debate is missing a very important question. We should stop worrying about whether [specific product] is net harmful and start asking how we can fix the root cause of the problem by getting effective treatment to people with addictive personality disorder. And yes, inventing effective treatment and rolling it out on a population scale is a much harder problem than just banning [target of latest moral panic]. But it’s the problem that society needs to solve.
1. ^
  In my parable the principal bans ‘useful’ things, while authorities responding to addictive behaviour usually want to ban entertainment. That isn’t a crux for me. Entertainment is intrinsically valuable, and banning it costs utility—potentially large amounts of utility when multiplied across an entire population.
2. ^
  Or more than one something else. People can be addicted to multiple things, I’m just eliding that for readability.
3. ^
  Zvi quotes research saying that legalizing sports betting resulted in a 28% increase in bankruptcies, which suggests it might be more financially harmful than whatever other addictions people had before, but that’s about as much as we can say.

Dumbledore's Army 22 Nov 2024 18:42 UTC
0 points
0
in reply to: dr_s’s comment on: Cost, Not Sacrifice
I agree with your first paragraph. I think the second is off-topic in a way that encourages readers, and possibly you yourself, to get mind-killed. Couldn’t you use a less controversial topic as an example? (Very nearly any topic is less controversial.) And did you really need to compound the problem by assigning motivations to other people whom you disagree with? That’s a really good way to start a flame war.

Dumbledore's Army 15 Sep 2024 16:57 UTC
11 points
1
on: The Great Data Integration Schlep
I spent eighteen months working for a quantitative hedge fund. So we were using financial data—that is accounts, stock prices, things that are inherently numerical. (Not like, say, defining employee satisfaction.) And we got the data from dedicated financial data vendors, the majority from a single large company, who had already spent lots of effort to standardise it and make it usable. We still spent a lot of time on data cleaning.

Dumbledore's Army 23 Mar 2024 10:27 UTC
14 points
4
in reply to: Seth Herd’s comment on: What does “autodidact” mean?
The education system also tells students which topics they should care about and think about. Designing a curriculum is a task all by itself, and if done well it can be exceptionally helpful. (As far as I can tell, most universities don’t do it well, but there are probably exceptions.)

A student who has never heard of, say, a Nash equilibrium isn’t spontaneously going to Google for it, but if it’s listed as a major topic in the game theory module of their economics course, then they will. And yes, it’s entirely plausible that, once students know what to google for, then they find that YouTube or Wikipedia are more helpful than their official lecture content. Telling people they need to Google for Nash equilibria is still a valuable function.

Dumbledore's Army 23 Mar 2024 10:24 UTC
5 points
2
on: What does “autodidact” mean?
As Richard Kennaway said, there are no essences of words. In addition to the points others have already made, I would add: Alice learns what the university tells her to. She follows a curriculum that someone else sets. Bob chooses his own curriculum. He himself decides what he wants to learn. In practice, that indicates a huge difference in their relative personalities, and it probably means that they end up learning different things.

While it’s certainly possible that Bob will choose a curriculum similar to a standard university course, most autodidacts end up picking a curriculum wildly different. Maybe the university’s standard chemistry course includes an introduction to medical drugs and biochemistry, and Bob already knows he doesn’t care about that so he can skip that part. Maybe the university’s standard course hardly mentions superconducting materials but Bob unilaterally decides to read everything about them and make that his main focus of study.

Dumbledore's Army 18 Mar 2024 19:42 UTC
56 points
36
on: The Worst Form Of Government (Except For Everything Else We’ve Tried)
The argument given relies on a potted history of the US. It doesn’t address the relative success of UK democracy—which even British constitutional scholars sometimes describe as an elective dictatorship that notoriously doesn’t give a veto to minorities. It doesn’t address the history of France, Germany, Italy, Canada, or any other large successful democracy, none of which use the US system, most of which aren’t even presidential,

If you want to make a point about US history, fine. If you want to talk about democracy, please try drawing from a sample size larger than one.

Dumbledore's Army 29 Jan 2024 18:02 UTC
2 points
0
on: How to write better?
I second GeneSmith’s suggestion to ask readers for feedback. Be aware that this is something of an imposition and that you’re asking people to spend time and energy critiquing what is currently not great writing. If possible, offer to trade—find some other people with similar problems and offer to critique their writing. For fiction, you can do this on CritiqueCircle but I don’t know of an organised equivalent for non-fiction.

The other thing you can do is to iterate. When you write something, say to yourself that you are writing the first draft of X. Then go away and do something else, come back to your writing later, and ask how you can edit it to make it better. You already described problems like using too many long sentences. So edit your work to remove them. If possible, aim to edit the day after writing—it helps if you can sleep on it. If you have time constraints, at least go away and get a cup of coffee or something in order to separate writing time from editing time.

Dumbledore's Army 1 Nov 2023 14:34 UTC
2 points
0
on: Forecasting Questions: What do you want to predict on AI?
First, I just wanted to say that this is an important question and thank you for getting people to produce concrete suggestions.

Disclaimer, I’m not a computer scientist so I’m approaching the question from the point of view of an economist. As such, I found it easier to come up with examples of bad regulation than good regulation.

Some possible categories of bad regulation:

1 It misses the point.
- Example: a regulation only focused on making sure that the AI can’t be made to say racist things, without doing anything to address extinction risk.
- Example: a regulation that requires AI-developers to employ ethics officers or risk management or similar without any requirement that they be effective. (Something similar to cyber-security today: the demand is that companies devote legible resources to addressing the problem, so they can’t be sued for negligence. The demand is not that the resources are used effectively to reduce societal risk.)
NB: I am implicitly assuming that a government which misses the point will pass bad regulation and then stop because they feel that they have now addressed ‘AI safety’. That is, passing bad legislation makes it less likely that good legislation is passed.

2 It creates bad incentives
- Example: from 2027 the government will cap maximum permissible compute for training at whatever the maximum used by that date was. Companies are incentivised to race to do the biggest training runs they can before that date
- Example: restrictions or taxes on compute apply to all AI companies unless they’re working on military or national security projects. Companies are incentivised to classify as much of their research as possible as military, meaning the research still goes ahead, but it’s now much harder for independent researchers to assess safety, because now it’s a military system with a security classification.
- Example: the regulation makes AI developers liable for harms caused by AI but makes an exception for open-source projects. There is now a financial incentive to make models open-source
3 It is intentionally accelerationist, without addressing safety
- A government that wants to encourage a Silicon Valley type cluster in its territory offers tax breaks for AI research over and above existing tax credits. Result: they are now paying people for going into capabilities research, so there is a lot more capabilities research
- Industrial policy, or supply chain friendshoring, that results in a lot of new semiconductor fabs being built (this is an explicit aim of America’s IRA). The result is a global glut of chip capacity, and training AI ends up a lot cheaper than in a free-market situation.

Dumbledore's Army 22 Oct 2023 15:53 UTC
13 points
6
on: AI Safety is Dropping the Ball on Clown Attacks, and Mind Control in General
Although clown attacks may seem mundane on their own, they are a case study proving that powerful human thought steering technologies have probably already been invented, deployed, and tested at scale by AI companies, and are reasonably likely to end up being weaponized against the entire AI safety community at some point in the next 10 years.
I agree that clown attacks seem to be possible. I accept a reasonably high probability (c70%) that someone has already done this deliberately—the wilful denigration of the Covid lab leak seems like a good candidate, as you describe. But I don’t see evidence is that deliberate clown attacks are widespread. And specifically, I don’t see evidence that these are being used by AI companies. (I suspect that most current uses are by governments.)
I think it’s fair to warn against the risk that clown attacks might be used against the AI-not-kill-everyone community, and that this might have already happened, but you need a lot more evidence before asserting that it has already happened. If anything, the opposite has occurred, as the CEOs of all major AI companies signed onto the declaration stating that AGI is a potential existential risk. I don’t have quantitative proof, but from reading a wide range of media across the last couple of years, I get the impression that the media and general public are increasingly persuaded that AGI is a real risk, and are mostly no longer deriding the AGI-concerned as being low-status crazy sci-fi people.