All of oge's Comments + Replies

Here's my lightning talk: I collect advice for learning how to play chess. The largest collection I've found is a slim book called "Rapid Chess Improvement" by Michael de la Maza.

Please send me the rarest advice you've heard of for improving one's chess ability.

I think in non-poly couples, having a mostly Builder or Entertaining relationship is a sign that one or both parties is having an affair.

This is because all the happy couples I know seem to have ~50:50 ratio of building/entertaining. And the unhappy couples have ~0:0. And I only know non-poly couples.

Thanks for pointing out what I think are the common side effects of popular drugs.

Thanks for the quote. The song has haunting lyrics

I enjoyed reading parts 1--6. Is there any chance you could discuss some of your conclusions sooner than 2023? I'd love to tie my feelings of gloom at humans-aligning-themselves-to-algorithms to some call-to-action...

5Justin Bullock2y
Thank you for the comment and for reading the sequence! I posted Chapter 7 Welcome to Analogia! (https://www.lesswrong.com/posts/PKeAzkKnbuwQeuGtJ/welcome-to-analogia-chapter-7) yesterday and updated the main sequence page just now to reflect that. I think this post starts to shed some light on ways of navigating this world of aligning humans to the interests of algorithms, but I doubt it will fully satisfy your desire for a call to action.  I think there are both macro policies and micro choices that can help. At the macro level,  there is an over accumulation of power and property by non-human intelligences (machine intelligences,  large organizations, and mass market production). The best guiding remedy here that I've found comes from Huxley. The idea is pretty straightforward in theory: spread the power and property around and in the direction away from these non-human intelligences and towards as many humans as possible. This seems to be the only reasonable cure to the organized lovelessness and its consequence of massive dehumanization. At the micro level, there is some practical advice in Chapter 7 that also originates with Huxley. The suggestion here is that to avoid being an algorithmically aligned human,  choose to live filled with love, intelligence, and freedom as your guide posts. Pragmatically, one must live in the present, here and now, to experience those things fully.  I hope this helps, but I'm not sure it will.  The final thing I'd add at this point is that I think there's something to reshaping our technological narratives around machine intelligence away from its current extractive and competitive logics and directed more generally towards humanizing and cooperative logics. The Erewhonians from Chapter 7 (found in Samuel Butler's Erewhon) have a more extreme remedy: stop technological evolution, turn it backwards. But short of global revolution, this seems like proposing that natural evolution should stop. I'll be editing these 7 chapters,

(I think) he thinks that managing/regulating/policing those corporations is the best that humans are willing to do.

So we cannot automate the discovery of unknown unknowns?

(BTW nice diagrams)

Seconding your comment. I wish OP had stated their definition of Consciousness to help clear the confusion in the various terms.

2Rafael Harth2y
"What it's like to be someone"; agree it should have been in the survey.

I feel an aversion to thinking about topics related in any way to my dying. No elevated heart rate, unless my person or my livelihood is threatened.

The only safe AGI software is libre AGI software? I buy that.

5Arcayer2y
Given: Closed source Artificial General Intelligence requires all involved parties to have no irreconcilable differences. Thence: The winner of a closed source race will inevitably be the party with the highest homogeneity times intelligence. Thence: Namely, the CCP. Given: Alignment is trivial. Thence: The resulting AI will be evil. Given: Alignment is difficult. Given: It's not in the CCP's character to care. Thence: Alignment will fail. Based on my model of reality, closed sourcing AI research approaches the most wrong and suicidal decisions possible (if you're not the CCP). Closed groups fracture easily. Secrets breed distrust, which in turn breeds greater secrecy and smaller shards. The solution to the inevitable high conflict environment is a single party wielding overwhelming power. Peace, freedom- civilization starts with trust. Simply building aligned AI is insufficient. Who it is aligned with is absolutely critical. Given this, to create civilized AI, the creator must create in a civilized manner.

How is your book club different?

5Tornus2y
The key thing is that it's low-commitment / low-guilt. I was inspired to start it by a friend who started a book club during the pandemic, fell catastrophically behind on the reading, and ultimately ended up ghosting her own book club. I've noticed that book clubs tend to become machines for making people feel guilty / overloaded, so I tried hard to avoid that. We do a book every 2 - 3 months, and the default expectation is that people won't attend unless that specific book is interesting to them. Shortly before the discussion, I send out a summary of the book (which was my motivation for writing this), so that people can attend and participate without needing to finish (or even start) the book. It's still a fairly new endeavor, but it seems to be working so far.

The world needs more poetry. Thanks, Elmer.

Answer by ogeDec 21, 202112

What is the day to day work like?

From reading this site it looks like the “heavy-hitters” mostly debate timelines, and possible architectures. There doesn’t seem to be too much concrete work for new researchers to do…

I think we'd like a summary of how the decisions were arrived at

Yeah, and let's not build a machine that can lie very well.

1SD Marlow2y
This is a relevant point: An AI that can craft some misdirection into a game or story is showing a deeper level of understanding, but as it's within a context (game/story), that isn't really a lie. The question for MIRI is, does that kind of "knowledge about misdirection" serve as a dual-use technology, where said ability could be used in other circumstances?  

What if we use the commentary from chess games as thoughts?

The problem with commentary not made by the players themselves is that, as far as I understand it, the project wants the general thoughts of the player and not just the motivations for every specific move. Like, ideally, they want some stream of consciousness commentary style "oh look, that guy looks kind of tough, I'll go see if I can agro him. Oh no! he's way too strong... lets go hide behind this tree it looks kinda safe [...]". That's why I suggested the lets plays and not e-sports in general.

If they're ok with just noise-free motivational analysis, anything with good commentators might work, and chess is indeed a pretty clean option. 

6JJ Hepburn2y
Could do Go, Poker or some E-Sports with commentary. Poker unlike chess has the advantage that the commentators can see all of the players hands but the players can only see their own. Commentators often will talk about what a player must be thinking in this situation and account for what is observable to the player or not. This would certainly be easier to scale but not as good quality.

Signal-boosting this. Here's to more teams working together to get this bounty! ᾔ2

FYI the Faulkner annotated screenplays have about 3 sentences of annotation for every 10 pages.

link text

It'd be hard for humans to compete with AI unless humans can communicate with the AI in reasonable-sized chunks e.g. a 100-page document. Me, I think we should chat in 10-page documents or less ᾓ7ἿE‍♀️.

It'd be nice to have even a sentence-long summary at the end. There are interesting nuggets but I can't find a justified call-to-action.

1M. Y. Zuo2y
It’s a topic that likely doesn’t have any realistic call to action, since most of the factors are not within anyone’s control. But perhaps there is and if someone would like to share then feel free. It was crossposted from my blog to see if anyone had similar thoughts, a wider circulation then that is unnecessary for my intentions.

Props for emphasizing that a lot of these projects are “tech bundled with services”.

I believe that all software projects are in fact, “tech bundled with services” since the software will need upkeep in order to stay relevant to its users. Of course, I'm biased since I'd love to own one such project.

Thanks for asking this question. If you do start the reading programme, I hope you'll post your plans and progress so I can follow along.

Grinnellian, wut wut. [nnadioge] '06 on plans.

Answer by ogeSep 27, 20212

Off the top of my head: perhaps interact with a clique of people for a pre-determined time like a year, then ask them in multiple ways (anonymous feedback form, face-to-face convos etc.) what your impact was.

1Fer32dwt34r3dfsz2y
I was thinking about something along these same lines: providing anonymous surveys to my friends to have them gauge how their interactions with me affect their productivity, long and short-term contentment, emotional resilience, etc...  People generally seem to be poor estimators, so I might have to be somewhat weary of their responses. More quantitatively, I could record the duration of my meetings with people, the words and tones exchanged between me and others (the former of these would be easier if the interaction occurred digitally), and things like (roughly) how much exercise (esp. in terms of walking) they did as a result of meeting with me (there are several friends I have who only really exercise if I suggest it or offer to do it with them). Once the amount of exercise is measured, I could approximate how much well-being it confers them by referencing literature on the effects of exercise on mood and well-being.  The main problem here is that this is a lot of work, and I am unsure how meticulous I should be. Perhaps it is necessary to be as detailed as possible, i.e. to consider as many variables or features of well-being as possible, to quantify your first-order psychological effects on others.  The objective of an impact report might be able to offer me some direction. At this present moment, developing a human impact report for myself would be an interesting and illuminating project, and would likely generate more ideas for how measuring impact can be done better, or how the results of measuring impact can be understood or framed. If I operate with the knowledge that I am not optimizing for completeness, then I think that I can produce a report more quickly, and accelerate progress on this front. 

Just out of curiosity: what kinds of binaries do you need to reverse-engineer on a regular basis?

Bryan Caplan has been creating his "economics graphic novels" using an old "comic creator" software. He has a valid license, but they company that makes it went out decades ago, and the license server no longer exists. So I disabled the license-server check for him.

When I worked in mobile, I did it frequently. Customer would call us and say our SDK isn't working. I'd download their app off the app store, decompile it, and figure out exactly how they're using us.

It's also surprisingly frequent how often I want to step through a library or program that I'm u... (read more)

Could you please put the takeaways at the top of the article, as a tldr?

A tl;dr would make sense if this post were about a less sensitive subject. I feel the behaviorist techniques of operant conditioning are easy to misapply. They must be employed with altruistic compassion and respect for others' values. In this post, I emphasized the human side of things by putting the story first and the behaviorist lens last.

0Ruby3y
I second this. Solid post otherwise!

I don't know. Alcor offered to send my agents the post-mortem CT scan, but haven't gotten back about biopsies.

I think we agree that a safe audit would be desirable. We differ in thinking that the toxicity of ASC is a dealbreaker. Is this accurate?

The most detailed studies of the brain today, which show the locations of dozens of memory-related proteins, are done using aldehydes (the A in ASC) so I hope to be revived as an emulation; I have little hope that my physical body can be rewarmed as is partly because of the difficulty of getting cryoprotectant to every part of the brain.

Come to think of it, if you could be preserved pre-mortem in a territory where that is... (read more)

Cryopreservation causes lots of damage, always. What would this show?

Cryopreservation doesn't have to cause damage. For instance, Aldehyde Stabilized Cryopreservation (on pigs) doesn't https://doi.org/10.1016/j.cryobiol.2015.09.003

Brain biopsies, especially by cryo staff, sound dangerous.

Brain biopsies are performed in hospitals e.g. during brain cancer diagnostics. They should not be dangerous to perform

Is there any indication that cryo companies would comply with this? What would associated costs be?

I don't know.

5Josh Jacobson3y
1. I don't think this is done at any of the main cryonics organizations, right? Their methods are damaging in perhaps less predictable ways than this mechanism. 2. I think the statement is deceptive and I wouldn't want it being shared without further context. I had a conversation with my expert-friend about this method and the type of damage it causes. Quoting some parts of my text message conversation with this expert friend who wishes to be anonymous: Regarding your second response: The first paper I looked at on brain biopsies (https://link.springer.com/article/10.1007/s10143-019-01234-w) says: That sounds like a high risk of being very damaging to me, and that's from one biopsy by expert medical staff vs. your proposed multiple by non-experts.

Seconding the point to, save myself and then export the program thereof to the rest of the world.

One model for choosing good names:

(1) selecting the concepts to include in the name, (2) choosing the words to represent each concept, and (3) constructing a name using these words.

"How Developers Choose Names" (2021) by Feitelson et al. https://arxiv.org/abs/2103.07487

I have a technique for naming a thing. It goes like this. First, I realize that I can't find a good name, so I ask someone what to name it. But they don't understand what it is, so I describe it in more detail, and then notice that my description has the ideal name sitting in it.

In theory you could avoid the bit where you bother someone, by trying to describe it beforehand.

Hey Adele, could you please summarize (or link to) the exercises that helped you? I'm having wrist problems as well and I've found that cycling through a wide range of exercises helps

8Adele Lopez3y
Sure! There's not really a specific exercise as much as a method for dealing with issues. The following is just my interpretation of what he taught me in person, so YMMV. If you move your hand/wrist/arm around, you can usually identify a configuration where there's a threshold at which it "catches" and starts hurting noticeably more. Then you keep your arm fixed at that position, and make small motions along each of your wrist/elbow/shoulder joints one at a time, to identify where the issue is: * wrist movement crosses threshold --> issue is in your forearm * elbow movement crosses threshold --> issue is in your upper arm * shoulder movement crosses threshold --> issue is in your back or neck (typically near the shoulder blades or up your armpits) You should also be able to feel what side the pain is on, and so (still keeping it at the threshold configuration) you gently press in that area with your other hand until you identify an especially tender or "knotted" spot (it doesn't always hurt, I often have to go by feel). Then you try to put a shear force on this area, as if you're trying to move the overlying muscle or tendon out of the way for whatever is below. While holding it out of the way, you try slowly moving your arm past the threshold again. If successful, it will be further out than before (and if not, you probably have the spot a little off, or aren't putting the right kind of force on the area, so experiment a bit to find what works). You'll want to keep moving your arm back and forth across the former threshold, making sure to move all three joints in the process. Then you can release the force and ensure that the threshold remains at the new location (and if not, reapply the force and move through it more). It can help to have a partner to apply the force, since it is sometimes is required in really hard to reach locations. By the time you're having serious wrist issues, you likely have issues in several locations all along your arm, back and n

Hey Raj, I don't know you and you may DM me to have a conversation after which I can provide feedback.

It looks like Singer subscribes to the maxim, "Give until you're unable to take care of yourself and your loved ones."

I like this framing better than the drowning child analogy since the trigger for the action pattern of giving is well-defined. Thanks for pointing this article out, Chris_Leong.

6Chris_Leong3y
He suggests that he subscribes to such a maximum, but also proposes a weaker maxim - that we should give unless the bad that would happen to us is comparable in moral significance to the harm we would prevent (note: this says comparable in moral significance, not equal in morally significance).

Props for writing up these out-of-the-box ideas. Hemispherectomies especially. I wonder how risky the procedure is.

Haha, I live in Lagos too. One day we'll have a proper meetup :)

0DragonGod7y
Okay. :)

Where in Nigeria?

0DragonGod7y
I live in Lagos, but I'm currently at my University; it's in Ogun state. You're (presumably) Nigerian as well (Yoruba from your name)?

Malnutrition is the visible surface symptom of "these are uncivilised, backwards people caught in a series of petty tribal wars".

I agree.

Could you tell me how you came about the list of African backward values? I currently live in an African country; I'd like the names of all the values I'd need to instil to avoid seeing preventable suffering around me.

(FYI I'd thought that having a public list of salaries and paying higher taxes, a la Norway, would be mostly sufficient to fix things)

0The_Jaded_One7y
Thanks for your comment! Can you say which country? Not in particular, the human brain tends to collect overall impressions rather than keep track of sources. This sounds like a seriously tough battle.

Perhaps put these files up on the Internet Archive (http://archive.org/upload/) so they're preserved "forever"

0Deku-shrub7y
I know the guy running the githug repo and his sources, that'll do me for now.

Hey, turchin, do you mind explaining how you came about your final goals i.e. infinite evolution?

I'm looking for a way to test which final goal is more right. My current best guess for my final goal is, "avoiding pain and promoting play" and I've heard someone say, alternatively "beauty in the universe and eyes to see it." It would be neat if these different goals are reconcilable in some way.

1turchin7y
At the beginning, I should note that any goal which is not including immortality is stupid, as infinite existence will include realisation of almost all other goals. So immortality seems to be a good proxy for the best goal. It is better goal than pleasure. Pleasure is always temporary, and somewhat not interesting. However, there is something bigger than immortality. I call it "to become a God". But I can't just jump, or become enlightened, or whatever, it will be not me. I want to go all the way from now to infinitely complex, eternal and superintelligent and benevolent being. I think it is the most interesting way live. But it is not just fun. It is the meaning of life. And the "meaning" is what makes you work, even if there are no fun ahead. For example, if you care about the survival of your family, it gives you meaning. Or, speaking better, the meaning takes you. The idea of infinite evolution is also a meaning for the following reasons. There is a basic drive to evolve in every living being. When you choose a beautiful goal, you want to put your genes in the best possible place and create best possible children, and this is a drive that moves evolution. (Not very scientific claim, as sexual selection is not as well accepted as natural selection. So it is more like poetic expression of my feeling about natural drive to evolution). If one educate himself, read, travel etc, it all is parts of this desire for evolution. Even the first AI will immediately find it and start to self-improve. The desire to evolve is something like Nietzschen "will to power". But this will is oriented on the infinite space of future possible mind states. I would add that I spent years working in the theory of happiness. I abandoned it and I feel much better. I don't need to be happy, I just need to be in working condition to move to my mission: infinitely evolve (it also includes saving humanity from x-risks and giving life extension for all, so my goal is altruistic). It may lo

I would love to hear your opinion since I have many loved ones currently planning to have kids. Do you mind DMing me your opinion?

Hi SquirrellInHell, how would you respond to the comment left on the original post that having kids is likely net bad for the kids but net good for the world?

Instead, the process you describe of "overthinking your motivations" and acting on abstract reasoning has led me to believe that having kids actually harms the kids. I'm not sacrificing their lives to the altruism of letting them improve the lives of everyone else.

0SquirrelInHell7y
I disagree, but I would hate to press my opinion on this in any way

Hey Duncan, where can I sign up for this?

2[DEACTIVATED] Duncan Sabien7y
Huh, I had not imagined Oge alongside the other probable housemates, but I'm imagining it now and it's certainly interesting. Send me an email with more details about your interest?

Hey Ray, would you mind posting the notes from the unconference? With the CFAR hackathon coming up, the notes might give me ideas of hacks to work on.

3Raemon7y
Finally did so. :) http://lesswrong.com/r/discussion/lw/p1f/notes_from_the_hufflepuff_unconference_part_1/
0Raemon7y
Yes, will do soon. I kept putting it off because I wanted to do a good summary of said notes but kept not having time. But I'll err on the side of posting something soon

Thanks for writing this Gordon. I was struck by this paragraph:

By making clever choices in the company we keep and the cultures we engage, as adults we can insulate ourselves from the fullness of the world

I'd never considered that my cleverness in avoiding aversive stimuli could hurt me in the long run.

Haha, I'd have advised gworley to drop everything before the 8th paragraph ("It’s in this spirit that I advise you, act into fear. ") since it could have been summarized by a paragraph-long disclaimer.

Lovely story, Stuart. I like how you captured certain aspects of thinking I hadn't seen articulated before e.g.

  • a being moving their attention/power towards different abilities as needed
  • that feeling of slowly becoming the part of a community
  • the feeling of noticing that a statement is likely not true
Load More