LESSWRONG
LW

RobertM
4597Ω4117048069
Message
Dialogue
Subscribe

LessWrong dev & admin as of July 5th, 2022.

Posts

Sorted by New

Wikitag Contributions

Comments

Sorted by
Newest
6RobertM's Shortform
2y
90
Don't Eat Honey
RobertM1d2919

If we assume conservatively that a bee’s life is 10% as unpleasant as chicken life

This doesn't seem at all conservative based on your description of how honey bees are treated, which reads like it was selecting for the worst possible things you could find plausible citations for.  In fact, very little of your description makes an argument about how much we should expect such bees to be suffering in an ongoing way day-to-day.  What I know of how broiler chickens are treated makes suffering ratios like 0.1% (rather than 10%) seem reasonable to me.  This also neglects the quantities that people are likely to consume, which could trivially vary by 3 OoM.

If you're a vegan I think there are a bunch of good reasons not to make exceptions for honey.  If you're trying to convince non-vegans who want to cheaply reducing their own contributions to animal suffering, I don't think they should find this post very convincing.

Reply111
TurnTrout's shortform feed
RobertM2d2519

I agree it's more related than a randomly selected Nate post would be, but the comment itself did not seem particularly aimed at arguing that Nate's advice was bad or that following it would have undesirable consequences[1].  (I think the comments it was responding to were pretty borderline here.)

I think I am comfortable arguing that it would be bad if every post that Nate made on subjects like "how to communicate with people about AI x-risk" included people leaving comments with argument-free pointers to past Nate-drama.

The most recent post by Nate seemed good to me; I think its advice was more-than-sufficiently hedged and do not think that people moving in that direction on the margin would be bad for the world.  If people think otherwise they should say so, and if they want to use Nate's interpersonal foibles as evidence that the advice is bad that's fine, though (obviously) I don't expect I'd find such arguments very convincing.

  1. ^

    When keeping in mind its target audience.

Reply
TurnTrout's shortform feed
RobertM2d2421

I think it would be bad for every single post that Nate publishes on maybe-sorta-related subjects to turn into a platform for relitigating his past behavior[1].  This would predictably eat dozens of hours of time across a bunch of people.  If you think Nate's advice is bad, maybe because you think that people following it risk behaving more like Nate (in the negative ways that you experienced), then I think you should make an argument to that effect directly, which seems more likely to accomplish (what I think is) your goal.

  1. ^

    Which, not having previously expressed an opinion on, I'll say once - sounds bad to me.

Reply1
Racial Dating Preferences and Sexual Racism
RobertM8d42

(Separately, even accepting for the sake of argument that you notice most work done and have a negative reaction to it, that is not very strong counterevidence to the original claim.)

Reply
Racial Dating Preferences and Sexual Racism
RobertM8d32

If the only thing you see about Aella is that she had work done on her lips, then I think that sufficiently well demonstrates the point that you don't notice most "high quality" plastic surgery.

Reply
Comparing risk from internally-deployed AI to insider and outsider threats from humans
RobertM8dΩ583

They imagine writing small and carefully locked-down infrastructure and allowing the AIs to interact with it.

That's surprising and concerning.  As you say, if these companies expect their AIs to do end-to-end engineering and R&D tasks internally, it seems difficult to imagine how they could do that without having employee-level privileges.  Any place where they don't is a place where humans turn into a bottleneck.  I can imagine a few possible objections to this:

  1. They don't expect those bottlenecks to impose that much of a penalty.
    1. I'm not sure how this could be true unless you think that AI systems will be pretty weak, which is sort of fighting the premise.  What are we worried about, again?
  2. They expect the bottlenecks to impose a large penalty and plan on biting that bullet when the time comes.
    1. I currently roll to disbelieve based on both the publicly-held positions of the relevant organizations and also their historical track records.  Incentives here seem quite bad.
  3. They plan on trying to thread the needle by employing some control schemes where (for example) different "agents" have different permissions.  i.e. a "code writing" agent has read permissions for (some parts of) the codebase, the ability to write, deploy, and test changes to that code in a sandboxed dev environment, and the ability to open a pull request with those changes.  Another set of agents have permissions to review pull requests, and then request changes, approve/merge them, or flag the PR as suspicious.  Yet another set of agents act as gatekeepers to sensitive data that might be needed for some experiments but only with good justification.
    1. This still suffers from the incentive gradient pushing quite hard to just build end-to-end agents.  Not only will it probably work better, but it'll be straight up cheaper and easier!

Like, to be clear, I would definitely prefer a world where these organizations wrote "small and carefully locked-down infrastructure" as the limited surface their AIs were allowed to interact with; I just don't expect that to actually happen in practice.

Reply
How does the LessWrong team generate the website illustrations?
Answer by RobertMJun 23, 2025210

This comment describes how the images for the "Best of LessWrong" (review winners) were generated.  (The exact workflow has varied a lot over time, as image models have changed quite a lot, and LLMs didn't always exist, and we've built more tooling for ourselves, etc.)

The prompt usually asks for an aquarelle painting, often in the style of Thomas Schaller.  (Many other details, but I'm not the one usually doing artwork, so not the best positioned to point to common threads.)  And then there's a pretty huge amount of iteration and sometimes post-processing/tweaking.

Reply2
Said Achmiz's Shortform
RobertM10d1010

Almost every comment rate limit stricter than "once per hour" is in fact conditional in some way on the user's karma, and above 500 karma you can't even be (automatically) restricted to less than one comment per day:

https://github.com/ForumMagnum/ForumMagnum/blob/master/packages/lesswrong/lib/rateLimits/constants.ts#L108

  // 3 comments per day rate limits
    {
      ...timeframe('3 Comments per 1 days'),
      appliesToOwnPosts: false,
      rateLimitType: "newUserDefault",
      isActive: user => (user.karma < 5),
      rateLimitMessage: `Users with less than 5 karma can write up to 3 comments a day.<br/>${lwDefaultMessage}`,
    }, 
    {
      ...timeframe('3 Comments per 1 days'), // semi-established users can make up to 20 posts/comments without getting upvoted, before hitting a 3/day comment rate limit
      appliesToOwnPosts: false,
      isActive: (user, features) => (
        user.karma < 2000 && 
        features.last20Karma < 1
      ),  // requires 1 weak upvote from a 1000+ karma user, or two new user upvotes, but at 2000+ karma I trust you more to go on long conversations
      rateLimitMessage: `You've recently posted a lot without getting upvoted. Users are limited to 3 comments/day unless their last ${RECENT_CONTENT_COUNT} posts/comments have at least 2+ net-karma.<br/>${lwDefaultMessage}`,
    }, 
  // 1 comment per day rate limits
    {
      ...timeframe('1 Comments per 1 days'),
      appliesToOwnPosts: false,
      isActive: user => (user.karma < -2),
      rateLimitMessage: `Users with less than -2 karma can write up to 1 comment per day.<br/>${lwDefaultMessage}`
    }, 
    {
      ...timeframe('1 Comments per 1 days'),

      appliesToOwnPosts: false,
      isActive: (user, features) => (
        features.last20Karma < -5 && 
        features.downvoterCount >= (user.karma < 2000 ? 4 : 7)
      ), // at 2000+ karma, I think your downvotes are more likely to be from people who disagree with you, rather than from people who think you're a troll
      rateLimitMessage: `Users with less than -5 karma on recent posts/comments can write up to 1 comment per day.<br/>${lwDefaultMessage}`
    }, 
  // 1 comment per 3 days rate limits
    {
      ...timeframe('1 Comments per 3 days'),
      appliesToOwnPosts: false,
      isActive: (user, features) => (
        user.karma < 500 &&
        features.last20Karma < -15 && 
        features.downvoterCount >= 5
      ),
      rateLimitMessage: `Users with less than -15 karma on recent posts/comments can write up to 1 comment every 3 days. ${lwDefaultMessage}`
    }, 
  // 1 comment per week rate limits
    {
      ...timeframe('1 Comments per 1 weeks'),
      appliesToOwnPosts: false,
      isActive: (user, features) => (
        user.karma < 0 && 
        features.last20Karma < -1 && 
        features.lastMonthDownvoterCount >= 5 &&
        features.lastMonthKarma <= -30
      ),
      // Added as a hedge against someone with positive karma coming back after some period of inactivity and immediately getting into an argument
      rateLimitMessage: `Users with -30 or less karma on recent posts/comments can write up to one comment per week. ${lwDefaultMessage}`
    },

I think you could make an argument that being rate limited to one comment per day is too strict given its conditions, but I don't particularly buy this as argument against rate limiting long-term commenters in general.

But presumably you want long-term commenters with large net-positive karma staying around and not be annoyed by the site UI by default.

A substantial design motivation behind the rate limits, beyond throttling newer users who haven't yet learned the ropes, was to reduce the incidence and blast radius of demon threads.  There might be other ways of accomplishing this, but it does require somehow discouraging or preventing users (even older, high-karma users) from contributing to them.  (I agree that it's reasonable to be annoyed by how the rate limits are currently communicated, which is a separate question from being annoyed at the rate limits existing at all.)

Reply
Documents Are Dead. Long Live the Conversational Proxy.
RobertM12dModerator Comment64

Hi Bharath, please read our policy on LLM writing before making future posts consisting almost entirely of LLM-written content.

Reply
Estrogen: A trip report
RobertM13d82

In a lot of modern science, top-line research outputs often look like "intervention X caused 7% change in metric Y, p <0.03" (with some confidence intervals that intersect 0%).  This kind of relatively gear-free model can be pathological when it turns out that metric Y was actually caused by five different things, only one of which was responsive to intervention X, but in that case the effect size was very large.  (A relatively well-known example is the case of peptic ulcers, where most common treatments would often have no effect, because the ulcers were often caused by an H. pylori infection.)

On the other end of the spectrum are individual trip reports self-experiments.  These too have their pathologies[1], but they are at least capable of providing the raw contact with reality which is necessary to narrow down the search space of plausible theories and discriminate between hypotheses.

With the caveat that I'm default-skeptical of how this generalizes (which the post also notes), such basic foundational science seems deeply undersupplied at this level of rigor.  Curated.

  1. ^

    Taking psychedelic experiences at face value, for instance.

Reply
Load More
73Briefly analyzing the 10-year moratorium amendment
1mo
1
31"The Urgency of Interpretability" (Dario Amodei)
2mo
23
207Eliezer's Lost Alignment Articles / The Arbital Sequence
4mo
10
281Arbital has been imported to LessWrong
4mo
30
29Corrigibility's Desirability is Timing-Sensitive
6mo
4
87Re: Anthropic's suggested SB-1047 amendments
1y
13
46Enriched tab is now the default LW Frontpage experience for logged-in users
1y
27
77[New Feature] Your Subscribed Feed
1y
13
30Against "argument from overhang risk"
1y
11
71LW Frontpage Experiments! (aka "Take the wheel, Shoggoth!")
1y
27
Load More
Simulation Argument
23d
(-1)
AI Safety & Entrepreneurship
1mo
Eliezer's Lost Alignment Articles / The Arbital Sequence
4mo
(+3/-4)
Solomonoff induction
4mo
Reinforcement learning
4mo
Probability theory
4mo
Probability theory
4mo
(-3)
Probability theory
4mo
(+56/-63)
Philosophy
4mo
Philosophy
4mo
Load More