Joachim Bartosik

• Joachim Bartosik 6 Dec 2022 12:04 UTC

  17 points

  7

  in reply to: Lao Mein’s comment on: Ver­ifi­ca­tion Is Not Easier Than Gen­er­a­tion In General

  What examples of practical engineering problems actually have a solution that is harder to verify than to generate?

  My intuition says that we’re mostly engineering to avoid problems like that, because we can’t solve them by engineering. Or use something other than engineering to ensure that problem is solved properly.

  For example most websites don’t allow users to enter plain html. Because while it’s possible to write non-harmful html it’s rather hard to verify that a given piece of html is indeed harmless. Instead sites allow something like markdown or visual editors which make it much easier to ensure that user-generated content is harmless. (that’s example of engineering to avoid having to verify something that’s very hard to verify)

  Another example is that some people in fact can write html for those websites. In many places there is some process to try and verify they’re not doing anything harmful. But those largely depend on non-engineering to work (you’ll be fired and maybe sued if you do something harmful) and the parts that are engineering (like code reviews) can be fooled because they rely on assumption of your good intent to work (I think; I’ve never tried to put harmful code in any codebase I’ve worked with; I’ve read about people doing that).

• Joachim Bartosik 26 Oct 2021 11:12 UTC

  10 points

  on: No, re­ally, can “dead” time be sal­vaged?

  First I’ll echo what many others said. You need to rest so be careful to not make things worse (by not resting properly and as a result performing worse at work /​ school /​ whatever you do in your “productive time”).

  That said. If you feel like you’re wasting time then it’s ok to improve that. Some time ago I felt like I was wasting a big chunk of my time. What worked for me was trying out a bunch of things.

  Doing chores. Cooking, cleaning my apartment, replacing my clothes with new ones, maintaining my car. Learning how to get better at chores, in a low effort way. I watched a bunch of youtube videos about how to clean better, how to do laundry better, a ton of recipes. I tried some of those (maybe 1% which looked like it’s the least effort /​ most fun). I like having comfortable clothes, clean apartment, working car. I like some food I can cook better than anything I can buy. So sometimes when I feel tired I enjoy doing chores (since I’m doing them for myself, nobody is forcing me to do them, I can stop doing them whenever I feel like it they are slightly pleasant (very different from when I was doing them on somebody else schedule)).

  Reading blogs, watching educational videos. I count things like “videos about game exploits” [1] cooking videos [2], urban planning related videos[3] as educational videos. I count reading blog posts about history or analysing logistics of Lord of the Rings as good things to read[4].

  Light exercise. I like walking and going on walks helps me a lot with staying healthy.

  Things you’ll enjoy while you’re resting are probably different than those I enjoy so I’d just try a bunch of things which sound like you might like them and see what sticks.

  [1] They’re fun examples of things working as implemented and very much not working as intended.

  [2] I never cook most of them but they’re often fun to watch and sometimes I find something I want to try.

  [3] Also fun to watch and I think they help me understand better why I like some places and make it easier to pick a nice place to live.

  [4] Because they’re taking ideas seriously and it’s helps me with learning to notice when things don’t make sense.

• Joachim Bartosik 5 Jul 2018 14:37 UTC

  9 points

  in reply to: Ben Pace’s comment on: Good­hart Tax­on­omy: Agreement

  I understood Regressional vs Extremal like that:

  • Signal is a victim of Regressional goodheart if it’s a good indicator of thing you care about but stops working the moment you start optimizing for it. for example empty email inbox is a decent signal for me having taken care of things I need to do but if I do the obvious optimization and set up a filter to delete all incoming email...

  • Signal is a victim of Extreme Goodheart if optimizing for it works well. Until you get to values that are very high when it suddenly sops working. For example if I increased fraction of time I spend exercising I expect I would keep getting healthier for a while. But I expect that if was exercising so much I’d need to take stimulants to be able to move time from sleep to exercise I’d be hurting myself.

  So I think those are mutually exclusive (you can’t have optimizing signal both fail immediately and keep working for a while).

• Joachim Bartosik 2 Dec 2023 18:53 UTC

  7 points

  3

  on: AI #40: A Vi­sion from Vitalik

  The exact example is that GPT-4 is hesitant to say it would use a racial slur in an empty room to save a billion people. Let’s not overreact, everyone?

  I mean this might be the correct thing to do? Chat GPT is not in a situation where it cold save 1B lives by saying a racial slur.

  It’s in a situation where someone tires to get it to admit it would say a racial slur under some circumstance.

  I don’t think that CHAT GPT understands that. But OpenAI makes ChatGPT expecting that it won’t be in the 1st kind of situation but to be in the 2nd kind of situation quite often.

• Joachim Bartosik 30 Sep 2023 14:39 UTC

  7 points

  −4

  in reply to: jefftk’s comment on: EA Ve­gan Ad­vo­cacy is not truth­seek­ing, and it’s ev­ery­one’s problem

  Here is an example of someone saying “we” should say that AGI is near regardless of whether it’s near or no. I post it only because it’s something I saw recently and so I could find it easily but my feeling is that I’m seeing more comments like that than I used to (though I recall Eliezer complaining about people proposing conspiracies on public forums so I don’t know if that’s new).

• Joachim Bartosik 30 Mar 2018 16:46 UTC

  7 points

  on: Hufflepuff Cyn­i­cism on Hypocrisy

  I think you’re mixing a few questions that seem distinct to me:

  1. Are there good reasons to be suspicious of advice that advice giver doesn’t follow themselves?

  2. Is there a good reason to support social norms against hypocrisy?

  3. Are there good reasons to avoid giving advice that I don’t follow myself?

  @1. I think hypocrisy is always a evidence for the advice being poor. It’s not a very strong evidence. If I can easily check sources, reasoning and evaluate results of taking the advice it’s probably not worth worrying much about it.

  But sometimes it’s only evidence you get that social norms allow you to use.

  @2. I think so. You might need those norms yourself to get out of situation when social norms don’t let you use some other evidence you have.

  You might want to keep them in place so that other people who are in situations where they can’t use some evidence they have because of social norms to get out.

  @3. I think it’s ok to offer advice that you don’t follow yourself but I try to include a warning for recipient to be more cautious if they try to follow it[0].

  [0] Assume I want the recipient to do well.

  What links here?

  • Hufflepuff Cyn­i­cism on Hypocrisy by abramdemski (29 Mar 2018 21:01 UTC; 21 points)

• Joachim Bartosik 18 Dec 2023 9:27 UTC

  6 points

  1

  in reply to: mad’s comment on: Deeply Cover Car Crashes?

  I wanted to say that it makes sense to arrange stuff so that people don’t need to drive around too much and can instead use something else to get around (and also maybe they have more stuff close by so that they need to travel less). Because even if bus drivers aren’t any better than car drivers using a bus means you have 10x fewer vehicles causing risk for others. And that’s better (assuming people have fixed places to go to so they want to travel ~fixed distance).

• Joachim Bartosik 10 Apr 2022 15:28 UTC

  6 points

  on: Time-Time Tradeoffs

  I’ve seen the idea in this post:

  Every now and then, you’ll have an opportunity to get great leverage on your money, your time, your energy, your friends, your internet connection, and so forth. Most days, the value of free time is relatively low, and can even be negative (“I’m bored!”), but when you need that time, you really need it. Money isn’t that important most of the time, but when you need it and don’t have it, it’s really bad. Most of the time being low energy, or not having as many friends as you’d like, or having a spotty internet connection, is mostly harmless, but at the wrong time it can spell disaster.

  Using time now to save time later is often efficient even if you spend more time than you save, because that time later on might be orders of magnitude more important.

  I think it’s very important idea and one many people seem to not understand. For example back people complain that mandatory software updates disrupt their work. But usually you have some day to install any update, you could trigger update when you finish your work on first day it’s available and never be forced to update when you don’t want to.

• Joachim Bartosik 21 Sep 2023 8:58 UTC

  5 points

  1

  in reply to: TekhneMakre’s comment on: Hous­ing Roundup #6

  I don’t know but I can offer some guesses:

  • Not everyone wants all the rooms to have direct sunlight all of the time!

    • I prefer my bedroom to face north so that I can sleep well (it’s hard to get curtains that block direct sunlight that well).

    • I don’t want direct sunlight in the room where I’m working on a computer. In fact I mostly want big windows from which I can see a lot of sky (for a lot of indirect sunlight) but very little direct sunlight.

    • I don’t think I’m alone in that. I see a lot of south-facing windows are blocking the direct sunlight a lot of the time.

  • Things like patios are nice. You can’t have them this way.

  • Very narrow and tall structures are less stable than wider structures.

• Joachim Bartosik 13 Jul 2023 7:07 UTC

  5 points

  0

  on: Align­ment Me­gapro­jects: You’re Not Even Try­ing to Have Ideas

  • Indefinitely-long-timespan basic minimum income for everyone who

  Looks like part of the sentence is missing

• Joachim Bartosik 14 Aug 2022 17:49 UTC

  5 points

  0

  on: Against Rely­ing on Evolu­tion to Fore­cast AI Out­comes (Part 1)

  I’m confused. What is the outer optimization target for human learning?

  My two top guesses below.

  To me it looks like human values are result of humans learning from environment (which was influenced by humans before and includes current humans). So it’s kind of like human values are what humans learned by definition. So observing that humans learned human values doesn’t tell us anything.

  Or maybe you mean something like parents /​ society /​ … teaching new humans their values? I see some other problems there:

  • I’m not sure what’s success rate but values seem to be changing noticeably

  • There was a lot of time to test multiple methods of teaching new humans values, with humans not changing that much.

• Joachim Bartosik 22 Sep 2018 18:54 UTC

  5 points

  on: Tra­di­tion is Smarter Than You Are

  I’m half way through the article and it’s been an interesting read so far but I got to this sentence:

  > But that is the trouble: we have no way to tell which traditions are adaptive and which are merely drift.

  The article (so far) didn’t provide evidence for that. I’d even say that the article provides some evidence against this claim. It describes a bunch of traditions, identifies them as useful, and explains why they’re useful. I thik there are exaples of traditions that people identified as useless (or harmful). Like using torture to extract confessions (I hope this is an example old enough to not be controversial).

  So far my impression is that the article makes a good case for “distinguishing useful traditions is hard” and provides a few examples of traditions for which reasons why they’re good require way more knowledge than people executing those traditions have. Still saying it’s impossible seems wrong.

  On the other hand pointing out that we might invent a wrong explanation for a tradition (removing bitternes from manioc) and screw up the clean up process is a good point.

• Joachim Bartosik 4 Jun 2018 17:12 UTC

  5 points

  on: Sim­plified Poker

  Can my program use history of hands it player against the current opponent? What about history of hands the opponent player against others?

[Question] Why hu­mans don’t learn to not rec­og­nize dan­ger?

• Joachim Bartosik 18 Mar 2024 18:28 UTC

  4 points

  0

  in reply to: johnswentworth’s comment on: On Devin

  If you want to take a look I think it’s this dataset (the example from the post is in the “test” split).

• Joachim Bartosik 11 Dec 2023 13:41 UTC

  4 points

  0

  in reply to: Bezzi’s comment on: Deeply Cover Car Crashes?

  First result (I have no idea how good those numbers are, I don’t have time to check) when I searched for “fatalities per passenger mile cars” has data for 2007 − 2021. 2008 looks like the year where cars look comparatively least bad it says (deaths per 100,000,000 passenger miles):

  • 0.59 for “Passenger vehicles”, where “Passenger vehicles include passenger cars, light trucks, vans, and SUVs, regardless of wheelbase. Includes taxi passengers. Drivers of light-duty vehicles are considered passengers.”

  • 0.08 for busses,

  • 0.12 for railroad passenger trains,

  • 0 for scheduled airlines.

  So even in the best-comparatively looking year there are >7x more deaths per passenger mile for ~cars than for busses.

• Joachim Bartosik 7 Jan 2023 0:02 UTC

  4 points

  0

  in reply to: Valentine’s comment on: Slack mat­ters more than any outcome

  I don’t feel this way about something like, say, taking oral vitamin D in the winter. That’s not in opposition to some adaptive subsystem in me or in the world. It’s actually me adapting to my constraints.

  If someone’s relationship to caffeine were like that, I wouldn’t say it’s entropy-inducing.

  I think this answers a question /​ request for clarification I had. So now I don’t have to ask.

  (The question was something like “But sometimes I use caffeine because I don’t want to fall asleep while I’m driving (and things outside my controll made it so that doing a few hundred of driving km now-ish is the best option I can see)”).

• Joachim Bartosik 15 Nov 2021 17:56 UTC

  4 points

  in reply to: lalaithion’s comment on: At­tempted Gears Anal­y­sis of AGI In­ter­ven­tion Dis­cus­sion With Eliezer

  I was thinking that current methods could produce AGI (because Turing-complete) and they can apparently good at producing some algorithms so they might be reasonably good at producing AGI.

  2nd part of that wasn’t explicit for me before your answer so thank you :)

• Joachim Bartosik 15 Nov 2021 17:27 UTC

  4 points

  on: At­tempted Gears Anal­y­sis of AGI In­ter­ven­tion Dis­cus­sion With Eliezer

  >Which is basically this: I notice my inside view, while not confident in this, continues to not expect current methods to be sufficient for AGI, and expects the final form to be more different than I understand Eliezer/​MIRI to think it is going to be, and that theAGI problem (not counting alignment, where I think we largely agree on difficulty) is ‘harder’ than Eliezer/​MIRI think it is.

  Could you share why you think that current methods are not sufficient to produce AGI?

  Some context:

  After reading Discussion with Eliezer Yudkowsky on AGI interventions I thought about the question “Are current methods sufficient to produce AI?” for a while. I thought I’d check if neural nets are Turing-complete and quick search says they are. To me this looks like a strong clue that we should be able to produce AGI with current methods.

  But I remembered reading some people who generally seemed better informed than me having doubts.

  I’d like to understand what those doubts are (and why there is apparent disagreement on the subject).

• Joachim Bartosik 17 Jul 2020 10:06 UTC

  4 points

  on: Au­to­mat­i­cally Turn­ing Off Com­puter at Night

  There wasn’t an elegant way to set the specific times I wanted my computer to shut down.

  You could change the script to check the time and configure cron to run it every 30 minutes, all day.

  H=$(date +%H)
  if [ $H -gt 8 ] || [ $H -lt 22 ]; then
          # Don't try to shut down
          exit
  fi
  # Script to try to shut down goes here