Comment by gworley on Quantitative Philosophy: Why Simulate Ideas Numerically? · 2019-04-18T21:15:28.201Z · score: 2 (1 votes) · LW · GW
I am not sure I follow. The bots indeed do end up clustering into 4 to 5 different clusters, where each cluster represents a certain convergent view. By "keeping the affinity score", do you mean they keep track of the past interactions, not just compare current views at each step? That would be an interesting improvement, adding memory to the model, but that would be, well, an improvement, not necessarily something you put into a toy model from the beginning. Maybe you mean something else? I'm confused.

Oh, this paragraph seems to suggest your model has a lot more going on that I got from reading this post. Maybe if I followed you links I would find some more details (sounded like they were just extra details that could be skipped)? I got the impression you found a function that has a shape illustrative of what you want and that was it, but this sounds like there's a lot more going on not described in the text of this post!

Comment by gworley on Quantitative Philosophy: Why Simulate Ideas Numerically? · 2019-04-17T02:17:51.437Z · score: 9 (1 votes) · LW · GW

It's unclear to me that the model you construct has much relationship to your tested idea, or at least any more than any other words or images you could have provided to illustrate the idea. I get that you see peaks and valleys in the function that you interpret as related to attraction and repulsion of people towards clusters, but again I don't see any clear grounding of this model in the idea you are testing, so to me it just feels like you drew a picture to illustrate the idea in a very convoluted way. Put another way, I don't feel like your model constrains your belief in the idea at all, because you could have come up with any function to draw any curve you could have wanted to show us.

Maybe I've missed something where you did provide a grounding, but I read back through the post a second time and still didn't see anything definitive. If there is something like that, or if you left it out and can explain how the model is causally connected to the idea such that something about the model says something about the idea, can you do so now?

FWIW I went in to this expecting a very different sort of model, one more like a simulation using simple bots that interact in simplified ways you describe and then we could see how they end up clustering, maybe by each bot keeping an affinity score for the others and finding results about the affinity of the bots forming clusters. That feels to me like it does have some more grounding in the idea being tested, where each bot is simplified stand-in for a person and their interactions and affinity scores for the other bots stand in for human interactions and human affinity for other humans. But again maybe you have some reason for thinking the model you present has the same kind of causal connection by sharing some structural similarity; I just don't see it so would appreciate if you could clarify.

Comment by gworley on Open Problems in Archipelago · 2019-04-16T23:59:23.690Z · score: 6 (4 votes) · LW · GW

Although I guess there's also the question of, why don't we just create an archipelago of subreddits on reddit if that's the direction we want to go? Just prepend the name of each subreddit with "LessWrong" and link them together somehow and be done with it.

I think we all know the answer, though: LW has certain standards and does a better job of keeping out certain kinds of noise than reddit does, even with active moderation. LW today attracts certain folks, deters others, and its boundaries make it a compelling garden to hang out in, even if not everyone agrees on, say, whether we should allow only flowering plants in our garden or if ferns and moss are okay.

I like the direction of having LW, EA Forum, and Alignment Forum being semi-connected; I would love if EA Forum functions more like the Alignment Forum does in relation to LW, and I think it would be cool to potentially see one or two additional sites branch off if that made sense, but I also don't feel like there's enough volume here that I'd enjoy seeing us fracture too much, because there's a lot of benefit too in keeping things together and exposing folks to things they otherwise might not see because it happens to be loosely connected enough to the things they do want to see that they end up encountering it. I enjoy stumbling on things I had no idea I would learn something from, but others are less open in this way and have different preferences.

Comment by gworley on Highlights from "Integral Spirituality" · 2019-04-16T19:10:00.293Z · score: 4 (2 votes) · LW · GW

The following quote was called out in a deleted comment, but I think there is something to discuss here that would be missed if we didn't come back to it even though that comment was ruled off-topic.

Thus, ultimate concern was displaced to science, a concern that its methods were simply not capable of handling. And science itself was always completely honest about its limitations: science cannot say whether God exists or does not exist; whether there is an Absolute or not; why we are here, what our ultimate nature is, and so on. Of course science can find no evidence for the Absolute; nor can it find evidence disproving an Absolute. When science is honest, it is thoroughly agnostic and thoroughly quiet on those ultimate questions.

The now deleted complaint was that this is saying something like science is in a non-overlapping magisterium from the question of whether or not God exists. I agree trying to claim separate magisterium is a problem and doesn't work, so what do I see as the value of including this quote?

Mainly to highlight a point that I think is often poorly understood: that science, for all the good it does, intentionally cuts itself off from certain kinds of evidence in order to allow it to function. Maybe we can debate what is the "real" science, but I'm thinking here of the normal, run-of-the-mill thing you'd call "science" we find going on in universities around the world, and that form of science specifically ignores lines of evidence we might call anecdotal or phenomenological and, for our purposes, ignores questions of epistemology by settling for a kind of epistemological pragmatism that allows science to get on with the business of science without having to resolve philosophy problems every time you want to publish a paper on fruit flies.

This choice to pragmatically ignore deep epistemological questions is a good choice for science, of course, because it lets it get things done, but it also means we cannot take results like "science finds no evidence of supernatural beings or some ever-present unifying force we could reasonably label God" as stronger evidence than it is. Yes, this is pretty strong evidence that there is no God like the kind you find in a religious text that interacts with the world, but it's also not much evidence of anything about a God that's more like an invisible dragon living in a garage. The thing that lets you address those sorts of questions is a bit different from what is typically done under the banner of science.

This quote does go a bit too far when it says science should be "thoroughly quiet on those ultimate questions", because it does have something to say, but I still thought it worth including because it highlights the common overreach of science into domains which it specifically rules itself our from participating in by setting up its methodological assumptions so that it can function.

(This last point put another way, think of how annoyed you'd be if every time you told your friend you felt sad and wanted a hug they said "I don't know, I can't really measure your sadness very well, and it's just you reporting this sadness anyway, so I can't tell if it's worth it to give you the hug".)

Comment by gworley on Slack Club · 2019-04-16T18:04:39.626Z · score: 3 (2 votes) · LW · GW

I have two thoughts on this.

One is that different spiritual traditions have their own deep, complex system of jargon that sometimes stretch back thousands of years through multiple translations, schisms, and acts of syncretism. So when you first encounter it you can feel like it's a lot and it's new and why can't these people just talk normally.

Of course, most LW readers live in a world full of jargon even before you add on the LW jargon, much of it from STEM disciplines. People from outside that cluster feel much the same way about STEM jargon as the average LW reader may feel about spiritual jargon. I point this out merely because I realized, when you brought up the spiritual example, that I wasn't given a full account of what's different about rationalists, maybe, in that there's a tendency to make new jargon even when a literature search would reveal existing jargon exists.

Which is relevant to your point and my second thought, which is that you are right, many things we might call "new age spirituality" have the exact same jargon-coining pattern in their writing as rationalist writing does, with nearly ever author striving to elevate some metaphor to the level of word so that it can becomes a part of a wider shared approach to ontology.

This actually seems to suggest then that my story is too specific and pointing to Eliezer's tendency to do this as a cause is maybe unfair: it may be a tendency that exists within many people, and there is something similar about the kind of people or the social incentives that are similar between rationalists and new age spiritualists that produces this behavior.

Comment by gworley on Slack Club · 2019-04-16T16:27:10.615Z · score: 14 (6 votes) · LW · GW
This archetype is easily distractible and does not cooperate with other instances of itself, so an entire community of people conforming to this archetype devolves into valuing abstraction and specialized jargon over solving problems.

Obviously there are exceptions to this, but as a first pass this seems pretty reasonable. For example, one thing I feel is going on with a lot of posts on LessWrong and posts in the rationalist diaspora is an attempt to write things the way Eliezer wrote them, specifically with a mind to creating new jargon to tag concepts.

My suspicion is that people see that Eliezer gained a lot of prestige via his writing, this is one of the things he does in his writing (name concepts with unusual names), and I suspect people make the (reasonable) assumption that if they do something similar maybe they will gain prestige from their writing targeted to other rationalists.

I don't have a lot of evidence to back this up, other than to say I've caught myself having the same temptation at times, and I've thought a bit about this common pattern I see in rationalist writing and tried to formulate a theory of why it happens that accounts not only for why we see it here but also why I don't see it as much in other writing communities.

Comment by gworley on Highlights from "Integral Spirituality" · 2019-04-15T21:54:46.737Z · score: 2 (1 votes) · LW · GW

I don't. Integral Spirituality might have some of what you're looking for, but only incidentally, since it's really trying to do something else.

Comment by gworley on The Meaning(s) of Life · 2019-04-14T16:59:58.566Z · score: 6 (3 votes) · LW · GW

Overall it seems like you're making a coherent enough point (meaningness is subjective), but I think the writing style makes that a bit hard to pick out. I can tell you've gotten some downvotes, and my guess is that it's because it's hard reading this to tell much about why you think this is or really given much in the way of specific arguments a person might engage with towards this point. I'm not saying I agree or disagree with you here, merely that I find it hard to follow your reasoning because it feels like there are many gaps in it in your writing, even if there aren't in your head.

I see you are rather new to the site, so I say this because I want to make sure you don't end up bouncing off because you get some downvotes. People generally respond well here to writing that is proof-like: well structured, aims to show something, and is clear about its starting assumptions (including assumptions about the audience; I felt like this post was arguing a point against an invisible other view that was never made very clear to me).

Hope that helps.

Highlights from "Integral Spirituality"

2019-04-12T18:19:06.560Z · score: 22 (19 votes)
Comment by gworley on Agent Foundation Foundations and the Rocket Alignment Problem · 2019-04-12T16:02:53.218Z · score: 2 (1 votes) · LW · GW

"Formally Stating the AI Alignment Problem" is probably the nicest introduction, but if you want a preprint of a more formal approach to how I think this matters (with a couple specific cases), you might like this preprint (though note I am working on getting this through to publication, have it halfway through review with a journal, and although I've been time constrained to make the reviewers' suggested changes, I suspect the final version of this paper will be more like what you are looking for).

Comment by gworley on Agent Foundation Foundations and the Rocket Alignment Problem · 2019-04-10T02:32:51.725Z · score: 2 (1 votes) · LW · GW

Both, although I mostly consider the former question settled (via a form of panpsychism that I point at in this post) and the latter less about the technical details of how AI could work and more about the philosophical predictions of what will likely be true of AI (mostly because it would be true of all complex, conscious things).

Also the "phenomenological" in the name sounded better to me than, say, "philosophical" or "continental" or something else, so don't get too hung up on it: it's mostly a marker to say something like "doing AI philosophy from a place that much resembles the philosophy of the folks who founded modern phenomenology", i.e. my philosophical lineage is more Kierkegaard, Hegel, Schopenhauer, Husserl, and Sartre than Hume, Whitehead, Russel, and Wittgenstein.

Comment by gworley on Agent Foundation Foundations and the Rocket Alignment Problem · 2019-04-09T21:55:23.789Z · score: 6 (3 votes) · LW · GW

I've been a bit busy with other things lately, but this is exactly the kind of thing I'm trying to do.

Comment by gworley on Value Learning is only Asymptotically Safe · 2019-04-08T19:02:21.296Z · score: 3 (2 votes) · LW · GW

I suspect there are many more sources of risk that result in only being able to approach complete safety than cosmic rays, but this seems a reasonable argument for at least establishing that the limit exists so even if we disagree over whether something more easily controlled by AI design is a source of risk we don't get confused and think that if we eliminate all risk from the design that we suddenly get perfect safety.

Comment by gworley on Ideas ahead of their time · 2019-04-04T20:15:57.517Z · score: 7 (5 votes) · LW · GW

This is a very outside view on these ideas. I think from the inside there's a lot that often separates obviously bogus ideas from possibly real ones. Ideas that might pan out are generally plausible now given the evidence available, even if they cannot be proved, whereas bogus, crank ideas generally ignore what we know to claim something contradictory. This can get a bit tricky because ideas of what people consider "known" can be a little fluid, but the distinction I'm trying to draw here is between ideas that may contradict existing models but agree with what we observe and ideas that disagree with what we observe (regardless of whether they contradict existing models), the former being plausible, ahead-of-their-time ideas that might later be proven true, and the latter being clearly bogus.

(Of course sometimes, as in the case of not observing star parallax without sufficiently powerful instruments, even our observations are a limiting factor, but this does at least allow us to make specific predictions that we should expect to see something if we had more powerful instruments, and would lead us to conclude against a promising idea if we got really good observations that generated disqualifying evidence.)

Comment by gworley on Coherence arguments do not imply goal-directed behavior · 2019-04-04T20:01:38.347Z · score: 2 (1 votes) · LW · GW

What I like about this thread, and why I'm worried about people reading this post and updating away from thinking that sufficiently powerful processes that don't look like what we think are dangerous is safe, is that it helps make clear that Rohin seems to be making an argument that hinges on leaky or even confused abstractions. I'm not sure any of the rest of us have much better abstractions to offer that aren't leaky, and I want to encourage what Rohin does in this post of thinking through the implications of the abstractions he's using to draw conclusions that are specific enough to be critiqued, because through a process like this we can get a clearer idea of where we have shared confusion and then work to resolve it.

Comment by gworley on Modelling Model Comparisons · 2019-04-04T18:05:02.311Z · score: 3 (2 votes) · LW · GW

Type 1 seems to be describing what I'd call a "structure" which is another way of talking about a pattern but in a certain abstract sense. For example, consider the classic mathematician joke about topologists not being able to distinguish a donut from a coffee cup because they have the same topological genus (at least, idealized donuts and coffee cups do), genus-1.

Type 2 seems to be describing what I'd call a "system", i.e. multiple objects in relation with each other coming together to form a new object at a different level of abstraction.

Although my thinking has certainly evolved a lot since then, I wrote about an issue that required addressing this topic a couple years go, so you might find that interesting even if you're not so interested in the topic I was addressing directly.

Comment by gworley on Ideas ahead of their time · 2019-04-04T17:44:09.982Z · score: 4 (3 votes) · LW · GW

I suspect panpsychism is in this boat. We have lots of philosophical reasons to think it makes sense, but making it the consensus requires overcoming two related difficulties:

  • convincing ourselves that "consciousness" is less special and magical than we currently think it is
  • reducing consciousness to something easily observable

Part of the problem seems to be we don't have the ability to adequately inspect the most complex conscious systems, and until we do it will remain possible to keep claiming "yeah, but real consciousness is special and not everything has it" because we imagine the simple pattern that strong theories of panpsychism propose explains consciousness is insufficient to explain the specialness of humans, animals, etc.

(This is not to be confused with weak theories of panpsychism, which are woo and reasonably dismissed (based on current evidence) because they propose the existence of phenomena we have not observed, like plants, rocks, and systems being as agentic as animals, but you know, in secret, or only on another plane of existence.)

Comment by gworley on Degrees of Freedom · 2019-04-04T00:33:59.101Z · score: 5 (2 votes) · LW · GW
this feels like it's moving things towards optimization and indifference.

I came here to say something like "I feel like this post sets up a false dichotomy" and I think you've done a better job than I would have at explicating why it feels to me optimization and indifference go together and are not really in opposition, except from within the prisons of our own minds thinking that they are in opposition.

Comment by gworley on What are effective strategies for mitigating the impact of acute sleep deprivation on cognition? · 2019-04-02T18:26:01.847Z · score: 2 (1 votes) · LW · GW

I don't have any literature on it, but it has had that effect on me. That is, as long as I'm meditating regularly (I average about 45 minutes a day of "serious" meditation, and another 60 minutes or so of "casual" meditation) I find if I don't get a full 10 hours I still often won't have sleep attacks (in fact I now normally only sleep about 8 hours most nights) and I can sleep as little as 6 hours and still function mostly normally (but not doing that repeatedly, and I will almost certain have a sleep attack on those days).

Comment by gworley on What are effective strategies for mitigating the impact of acute sleep deprivation on cognition? · 2019-04-01T18:07:59.475Z · score: 3 (2 votes) · LW · GW

An off-label use of fluoxetine (Prozac) is that it can caused prolonged sleep, possibly by reducing anxiety in ways that make it easier to stay asleep longer but specific mechanism of action is unknown. Worked well for me in treating narcolepsy-related sleep depravation, i.e allowed me to stay asleep 10 hours a night so I got enough sleep to avoid sleep attacks during the day. I'm no longer on it and still able to get enough sleep; my theory there is that regular meditation replaced the need for a drug to produce the same effect, allowing me to stay asleep longer.

Comment by gworley on Subagents, akrasia, and coherence in humans · 2019-03-31T21:30:07.178Z · score: 2 (1 votes) · LW · GW

Right, if you've not had the later experiences (equanimity, fruition leading to attainment) you're likely to mistake others for them, especially if you have a very squishy model of enlightenment and especially especially if you are trying hard to attain the path. My comment was more a reference to the fact that Ingram seems to view stream entry as a very precise thing relative to how it is talked about in theravada, which is why it seems possible that some of the above disagreement on numbers might be due to a different sense of what qualifies as stream entry.

I have my own fairly precise way of describing it, which is that you develop the capacity to always reason at Commons' MHC level 13 (this is placed about halfway along the 4 to 5 transition in the normal Kegan model by Wilburl but I consider that to be an inflation of what's really core 4), i.e. you S1 reason that way, deliberative S2 reasoning at that level is going to happen first but doesn't count. At least as of right now I think that, but I could probably be convinced to wiggle the location a little bit because I'm trying to project my internal model of it back out to other existing models that I can reference.

Comment by gworley on Subagents, akrasia, and coherence in humans · 2019-03-30T23:57:58.849Z · score: 4 (2 votes) · LW · GW

To be clear I am appropriating stream entry here the same way Ingram has, which has much more inclusive (because they are much smaller and more specific) criteria than what is traditional. I agree with your point about A&P, and maybe I am typical minding here because I made it through to what matches with once returner without formal practice (although I did engage in a lot of practices that were informal that drug me along the same way).

Comment by gworley on Timing through the temporal dimension. · 2019-03-30T23:50:01.026Z · score: 2 (1 votes) · LW · GW

You're just repeating yourself and still not offering an explanation of your theory, just some vague glimpses of it. I've tried twice now to ask specific questions that you could have answered to provide clarification, and rather than engage with them directly you chose either to presume I didn't read what you wrote or restate what you originally said in fewer words. LW it's about curiosity and inquiry among other things, and if that's not the spirit in which you've come here I won't continue to engage with you and will encourage others to do the same.

Comment by gworley on Timing through the temporal dimension. · 2019-03-29T19:20:05.290Z · score: 3 (2 votes) · LW · GW

Okay, so if I squint really hard maybe you are proposing a functional theory of time and just explaining it in a way that is not clear? For example, does this post about logical time comport with your model?

Saying "time is a verb" still doesn't say much, unfortunately, because lots of folks have known for quite some time that time as we normally think of it is an after-the-fact construction and not metaphysically basic. That is, it seems that our notion of time arises from how we perceive and remember events and is an expression of causality, that is whatever the fundamental way it is that the world changes from one state to another. If so, this is again not revolutionary, although maybe you are just unfamiliar; I can off the top of my head think of the likes of Dogen, Husserl, and Merleau-Ponty saying similar things, and I'm certain these ideas have a recorded exploration back at least 2000 where they were explored in Indian philosophy (though I can't remember the names attributed to those works now).

If that's not the case I'd really like to know, but you are giving us frustrating little to try to understand whatever you think your big idea is (hence, I suspect, the many downvotes you are receiving).

Parfit's Escape (Filk)

2019-03-29T02:31:42.981Z · score: 40 (15 votes)
Comment by gworley on Timing through the temporal dimension. · 2019-03-29T00:17:39.336Z · score: 2 (1 votes) · LW · GW
The word 'time' is a term describing temporal motion. So we 'move' through the spatial dimension and we 'time' through the temporal dimension.

You specifically call it a "temporal dimension".

The speed at which we time through the temporal dimension is called our temporal velocity.
We only time in one direction. Towards the future. So the future is a temporal location, towards which we 'time'. The past is a temporal location which we have already occupied.
Because the two dimensions are interlinked, our spatial velocity is inversely proportional to our temporal velocity. The faster we move, the slower we time. Timing, however, is not just for humans. Everything in the universe, besides mass less particles, times.

This further reads to me like you are exactly describing it as a dimension. If you mean something else you are not conveying it to me (and I doubt this is me being thick since I studied at least enough physics to be able to publish on quantum information theory). That last paragraph I quoted especially sounds like you are gesturing at relativity, a theory worked out over 100 years ago, which is why I asked what about your model is different, because I read your post and I can't tell.

Comment by gworley on What would you need to be motivated to answer "hard" LW questions? · 2019-03-28T21:53:01.142Z · score: 6 (3 votes) · LW · GW

Given that there is some probability of winning a question, let's just guess it's 20% on any particular question I might try to answer. This suggests to me a bounty of 5x whatever I would be willing to answer the question for in order to make me willing to do it. Assuming a question takes about a day of work (8 hours) to answer fully and successfully, and given our 5x multiplier, I'd be willing to try to answer a question I wasn't already excited to answer for other reasons if it paid about $1800.

Many others may have lower opportunity costs, though (and I undercounted a bit because I assume any question I would answer would deliver me at least some sense of value beyond the money; otherwise my number would probably jump up closer to $2500).

Comment by gworley on Timing through the temporal dimension. · 2019-03-28T19:32:50.179Z · score: 1 (2 votes) · LW · GW

This sounds very much to me like how time is dealt with in the standard model of physics, which leads to two questions. How, if at all, do you see this as different from the standard model (I'm guessing you think it is because you say "A completely new way of thinking about time." but I don't see any evidence of this)? And why then, if time is a dimension like this (and does not just, say, have a projection along a dimension that makes it easy to reason about in certain situations), does causality only flow in a single direction?

Comment by gworley on Model Mis-specification and Inverse Reinforcement Learning · 2019-03-28T18:05:07.551Z · score: 12 (3 votes) · LW · GW

Overall I think this piece is great and does a nice job of intuitively explaining ways our attempts to model human values can fail. I notice a bit of friction when I read this part, though:

How do we choose between the theory that Bob values smoking and the theory that he does not (but smokes anyway because of the powerful addiction)? Humans choose between these theories based on our experience with addictive behaviours and our insights into people’s preferences and values. This kind of insight can’t easily be captured as formal assumptions about a model, or even as a criterion about counterfactual generalization. (The theory that Bob values smoking does make accurate predictions across a wide range of counterfactuals.) Because of this, learning human values from IRL has a more profound kind of model mis-specification than the examples in Jacob’s previous post. Even in the limit of data generated from an infinite series of random counterfactual scenarios, standard IRL algorithms would not infer someone’s true values.

I see this kind of thing often in people's thinking: they intuitively have a sense that people can seem to value something and yet not value it because they don't endorse that value. I think this is a confused view, though, taken from our phenomenology of values and preferences when we are also identified with (subject to) those values and preferences and then, not liking what we see in ourselves, creating an ontology that suggests that some values and preferences is not endorsed, we would prefer for them to be otherwise, but find ourselves doing something we don't like anyway.

This sets up an interesting dialectic, because on one hand we have the very real, felt experience of wanting to do one thing (say, not smoke) and then doing the other (smoking anyway) and feeling as if doing the action (smoking) is not really what we want to do and not being "authentic" to our "true" or real self, and on the other we have the very real sense in which we are getting information about values and preferences based on behavior that suggests despite what we say ("I don't want to smoke") we don't act on it. Partly we might attribute this to a lack of reflective equilibrium resulting in irrational preference ordering, although I think that abstract away most of the interesting human psychology that produces this result. Anyway, I point this out because I think there is a useful synthesis that gets us beyond these two conflicting approaches that seem to get in our way of understanding human values: it's correct that we prefer, in this example, to smoke rather than not smoke, but it's also true that we believe we prefer to not smoke rather than smoke, and this is only a problem in that our model may be trying to assume that our preferences match our beliefs.

Now of course our beliefs can change our preferences, but that sounds a bit confusing if we just talk about beliefs and preferences because preferences would seem to be a special kind of belief relating to an ordering over actions, which I think shows that beliefs and preferences are a leaky abstraction. To resolve this we have to look a bit deeper, probably in the direction of Friston.

Comment by gworley on Subagents, akrasia, and coherence in humans · 2019-03-27T19:27:22.042Z · score: 3 (2 votes) · LW · GW

So I do think stream entry is way more common than most people would think because the thing that is stream entry is amazing and useful but also incredible normal and I think lots of folks are walking around having no idea they attained it (this relies, though, on a very parsimonious approach to what counts as stream entry). Whether my identifying it with Kegan 4 means the same thing as what that study from which that number comes does (which was itself, as I recall, not that great a study, and was lead by Lahey) is questionable since it depends on where you choose to draw the borders for each stage (the Subject-Object Interview manual provides one way of doing this, and is the method by which the number you mention was obtained).

My suspicion is that the number is much lower as I would count it, which I calculate as closer to 7% doing a Fermi estimate based on my own observations and other evidence I know of, even though a lot of folks (this is where I would say the 35% number makes sense) are somewhere in what I would consider the 3.5 to 4 range where they might be able to pass as 4 but have not yet had the important insight that would put them fully into the 4 stage.

So all those caveats aside, yes, I consider stream entry to be pointing at the same thing as Kegan 4.

Comment by gworley on Subagents, akrasia, and coherence in humans · 2019-03-26T20:49:05.284Z · score: 4 (2 votes) · LW · GW

No, I meant 3 to 4. What I think of as the 4 to 5 key insight builds on this one to say that not only can you think of yourself as a manipulable complex system/machinery and work with that, it takes a step back and says what you choose to make the system do is also able to be manipulated. That's of course a natural consequence of the first insight, but really believing it and knowing how to work with it takes time and constitutes the transition to another level because getting that insight requires the ability to intuitively work with an additional level of abstraction in your thinking.

Following on 5 to 6 is about stepping back from what you choose to make the system do and finding you can treat as object/manipulate how you choose (preferences; the system that does the choosing). Then 6 to 7 is about getting back one more level and seeing you can manipulate not just preferences but perceptions since they control the inputs that produce preferences.

Comment by gworley on Subagents, akrasia, and coherence in humans · 2019-03-26T20:41:17.946Z · score: 2 (1 votes) · LW · GW

I sort of started in this vicinity but then ended up somewhere else.

Comment by gworley on Do you like bullet points? · 2019-03-26T20:38:49.922Z · score: 2 (1 votes) · LW · GW
"I like numbered arguments, since that makes it easier to respond to individual points. But unnumbered bullet lists are just hard to parse."
[Alas, the LessWrong website currently doesn't enable this very well because our Rich Editor's implementation of numbered lists was annoying]

I don't much like writing in bullet points myself, but sometimes it's useful for particular purposes, and having good numbering, especially hierarchical numbering, would make being able to respond to bullet points really nice. For example, I once wrote a post (can't seem to find it now, so maybe it was a comment? closest example I could find was this) where I laid out numbered claims and conclusions to try to make it easier for people to engage with each one independently. Having numbering that looked like the following would be really nice:

1. top level

1.1 sub point

1.2 another sub point

1.2.1 sub sub point

2. another top level

Comment by gworley on Do you like bullet points? · 2019-03-26T20:31:26.890Z · score: 14 (4 votes) · LW · GW

This is tangental, but since you brought it up, I find the distinction you try to make between "explaining" and "persuading" weird, because what is "explaining" but persuading the reader to believe something, and the extent to which you've successfully explained something is the extent to which you've persuaded them to take up the same belief you have that the evidence and conclusions you draw from it are as you presented them. That is, all writing is persuasive writing.

The same is true if we try to reverse the situation, because what is persuading other than explaining something so well that the reader agrees with you, even if your "explanation" is unconventional in terms of what we often think of as explication.

Now there is some difference of intent of the author between what you might call a persuading-mode and an explaining-mode, I see you put "explaining" and "persuading" in quotes to denote your perhaps non-standard use, and perhaps you do this to imply you mean something more like what I'm suggesting in terms of authorial intent, but I suspect we can find more specific terms of the kinds of things that are in and out.

I bring this up because although I've never been called out for writing "persuasively" in the LW 2.0 era, I literally think of everything I write as a kind of persuasion—an attempt to say words that will cause the reader to have beliefs of roughly a certain kind after reading my words. And while I can safely ignore what you say about "explaining" and "persuading" and continue to contribute to the community because I have a rich model of what is and is not acceptable to the community, it likely pushes newer folks away from writing things that, on the margin, would be better if they were a little more persuasive because then they would be writing to get me to believe something rather than trying to actively avoid thinking about how the reader might respond to what's written and writing in response to what that model of the reader suggests needs to be written to get them to update.

I don't have a proposal for what words to use to describe the category of thing that is out, though; I've only gotten as far as noticing the dissonance but not sublimating it.

Comment by gworley on Subagents, akrasia, and coherence in humans · 2019-03-26T02:05:19.068Z · score: 11 (6 votes) · LW · GW
The original question which motivated this section was: why are we sometimes incapable of adopting a new habit or abandoning an old one, despite knowing that to be a good idea? And the answer is: because we don’t know that such a change would be a good idea. Rather, some subsystems think that it would be a good idea, but other subsystems remain unconvinced. Thus the system’s overall judgment is that the old behavior should be maintained.

To me this is the key insight for working with subagent models. Just to add something about the phenomenology of it, I think many people struggle with this because the conflicts can feel like failures to update on evidence, which feels like a failure as a result of identifying with a particular subagent (see a recent article I posted on akrasia that makes this same claim and tries to convince the reader of it in terms of dual-process theory). Thus this is a case of easily said, difficultly done, but I think just having this frame is extremely helpful for making progress because at least you have a way of thinking of yourself as not fighting against yourself but manipulating complex machinery that decides what you do.

As a bonus to my developmental psychology friends out there, I think this points to the key insight for making the Kegan 3 to 4 transition (and for my Buddhist friends out there, the insight that, once grokked, will produce stream entry), although your milage may vary.

Comment by gworley on Why the AI Alignment Problem Might be Unsolvable? · 2019-03-24T19:52:41.454Z · score: 9 (6 votes) · LW · GW

Ok, so a few things to say here.

First, I think the overall conclusion is right in a narrow sense: we can't solve AI alignment in that we can't develop an air tight proof that one agent's values are aligned with another's due to the epistemic gap between the experiences of the two agents where values are calculated. Put another way, values are insufficiently observable to be able to say whether two agents are aligned. See "Formally stating AI alignment" and "Outline of metarationality" for the models that lead me to draw this conclusion.

Second, as a consequence of what I've just described, I think any impossibility resulting from the irreconcilability of human preferences is irrelevant because the impossibility of perfect alignment is dominated by the above claim. I think you are right that we can't reconcile human preferences so long as they are irrational, and attempts at this are likely to result in repugnant outcomes if we hand the reconciled utility function to a maximizer. It's just less of a concern for alignment because we never get to the point where we could have failed because we couldn't solve this problem (although we could generate our own failure by trying to "solve" preference reconciliation without having solved alignment).

Third, I appreciate the attempt to put this in a fictional story that might be more widely read than technical material, but my expectation is that right now most of the people you need to convince of this point are more likely to engage with it through technical material than fiction, although I may be wrong about this so I appreciate the diversity of presentation styles.

Comment by gworley on Retrospective on a quantitative productivity logging attempt · 2019-03-22T15:30:28.690Z · score: 2 (1 votes) · LW · GW

I think it's somewhat a matter of personal taste, but like you I've found such attempts to quantify my life dissatisfying, although I know others who get a lot out of such attempts. I general fall in the direction of not bothering to measure hard to measure things if I don't have to, and when I'm reluctantly forced to do it I try to use very gross measurements to match the poor precision possible in such cases. Having the precision of the measurement match the level of precision you can achieve helps avoid getting confused by the numbers and thinking you have more information than you do.

Comment by gworley on Subagents, introspective awareness, and blending · 2019-03-13T02:27:40.798Z · score: 5 (2 votes) · LW · GW

This is giving me a new appreciation for why multiagent/subagent models of mind are so appealing. I used to think of them as people reifying many tiny phenomena into gross models that are easier to work with because the mental complexity of models with all the fine details is too hard to work with, and while I still think that's true, this gives me a deeper appreciation for what's going on, because it seems it's not just that it's a convenient model to abstract away from the details, but that the structure of our brains is setup in such a way that makes subagents feel natural so long as you model only so far down. I only ever much had a subjective experience of having two parts in my mind before blowing it all apart, so this gives an interesting look to me of how it is that people can feel like they are made up of many parts. Thanks!

Comment by gworley on The tech left behind · 2019-03-12T21:03:35.047Z · score: 13 (9 votes) · LW · GW

Oh, here's another one: Lisp Machines. These were computers with alternative chip designs focused on executing Lisp (or really any functional programming language) rather than on executing procedural code. Had the direction been pursued further, they might have resulted in dramatically different computer architectures than what we use today. Some were built and used, but only in very limited contexts, so I'd say this meets the criteria of "never saw the light of day" in that less than 10k Lisp machines were ever built.

[Old] Wayfinding series

2019-03-12T17:54:16.091Z · score: 9 (2 votes)

[Old] Mapmaking Series

2019-03-12T17:32:04.609Z · score: 9 (2 votes)
Comment by gworley on The tech left behind · 2019-03-12T17:10:00.298Z · score: 11 (5 votes) · LW · GW

One that comes to my mind is OpenDoc, a cool and exciting proposal for a way to make editable generic computer documents that were not application constrained. The idea was to make documents a cross-platform, operating system level responsibility and what we today think of as applications would instead be embedded viewers/editors that could be used when putting different types of "objects" in documents.

We did eventually get something like it: Google Docs, Word, and even web pages generally have the ability to embed all kinds of different other documents, and sometimes there is viewing/editing support within the document (you can see images, embed editable spreadsheets, embed editable diagrams, etc.), but with more vendor lock-in and missing the spirit of vendor openness OpenDoc intended.

Comment by gworley on If you wrote a letter to your future self every day, what would you put in it? · 2019-03-12T17:00:17.086Z · score: 4 (3 votes) · LW · GW

Well, I guess I sort of do this, although not with email and only for certain very narrow purposes: I use a few tools (calendars, emails to myself that describe an action item in the subject, a notes app that displays on my phone's home screen) to make sure I don't forget things I want to do something with later. This is basically just part of my personalization of the Getting Things Done method, although what I do these days doesn't look much like GTD as described in the books, but it's carried out in the same spirit: make a decision about what to do now, and then either do it now, put it in a system you trust so you are sure to do it later, or drop it.

Certainly an email to myself every day could achieve the same thing since I could edit it every day to contain the current state of the information I want to store in these systems, although I imagine more could be done with such a generalized mechanism that I can't do with my narrow way of using the tools I currently do.

Comment by gworley on Motivation: You Have to Win in the Moment · 2019-03-05T23:06:58.264Z · score: 0 (2 votes) · LW · GW
I do disagree with C (compelling only from a certain stage of development) in that I think even once you have much deeper understanding, the higher levels of abstraction remain crucially important. Just because you understand electromagnetism really well and know the limits of conventional circuit theory (e.g. designing super duper tiny transistors), doesn't mean you want to throw out circuit theory and just solve Maxwell's equations everywhere - even if eventually sometimes you have to.

So maybe it would help if I was a little more specific about this point. When I'm saying "compelling" here I meant to point to something like both intellectually interesting and useful because it feels new and like it's engaging with the edge of development. Stuff like this becomes uncompelling as one gains mastery, so I think I was trying to pass on the wisdom of my accumulated experience in this area from building, learning, using, and presenting models like this one and then, upon reconsidering, finding them limiting but having been useful at one point because I didn't have access to any deeper details to help me along.

My objective in pointing this out is tied in with the next bit, so we'll just go ahead and segue to that.

To be honest, I did bristle at some of the way things were phrased, but that's on me. It felt like there was some kind of implication that I personally didn't have any deeper understanding and that felt.

To be honest, there is an implication like that, based on what I've read here. I could maybe believe you intentionally didn't address some of the deeper points you might understand about the details that I think are relevant, but if that were the case I would expect your footnotes and asides to address topics more about beliefs, preferences, and especially perception and less about those things munged together and rounded off to "motivation". Instead I read this as your honest best effort to explain what's going on with motivation, and I'm telling you I think there's much more going on in directions much more fine-grained than those you seem to have explored, even in the references.

"Motivation" and "intention" are huge, confounded concepts that I believe can be broken apart, thinking of yourself as having a "motivation system" is another confusion, but unfortunately I've not worked out all the details well enough for myself that I'm happy to share my current state of partial knowledge in this area. Unfair, I admit, but it's where I stand. All I can point to is there's a bunch of stuff going on that can be reified into the concept of "motivation" and working with motivation as a concept will be helpful for a while but ultimately "motivation" doesn't cut reality at the joints so thinking in those terms has to be largely abandoned to go further.

Should I have publicly passed judgement on you in the comments section? Probably not, but for some reason I already did so we'll just have to deal with it now. Sorry about that.

My goal here is to be encouraging, however it might come across, and to make clear there is a way forward. As I said to another person recently when I responded in a similar way to something they said, I've been realizing a lot recently the ways in which I limited myself by thinking I understood things. I see in this work clues that you having an understanding similar to how I thought about motivation maybe 3 years ago, and maybe I would already have a ready-at-hand alternative if I hadn't spent so much time thinking I had it right. So I want you to explain what you've figured out, I think your way of explaining what you have is going to be useful for others, I don't want to say anything that might put you off either of those goals, and I also want to push you along so you don't suffer the worst of all calamities: thinking you understand something!

I also think D (unlikely to help many people) is somewhat false, depending on what counts as "many people". Another commenter felt this post was quite useful, someone else on FB found it rather revelationary, and I'd infer from those who I know of that several more benefited even if I don't know of it directly. That's beyond the inside view that abstraction/model presented can be applied already. mr-hire also states simpler ideas worked well for a really long time (though I'm not sure which simpler ideas or what counts as "brute force".

Sure, I guess I was hoping to set expectations appropriately, since I know I've been let down many times broaching these topics with folks. Yes, there will always be some people who you manage to connect with in part because of what you write and in part because of where they are, i.e. they are ready to listen to what you have to say and have it click. They are the cherished folks with little enough dust in their eyes that you write for. But for every person you help, there are probably 20 more who will read this and for one reason or another it won't connect the way you'd hope it would. They might not hate it, and might say they get it, but then they'll just keep on doing what they were doing, not changing anything really, not really having gained any understanding. I was demoralized a lot by this, thinking it must have been me, until I figured out the base rate of success for this kind of thing is pretty low unless you're tackling stuff way down at the bottom of the developmental ladder. I suspect, based on the quality of your explanation, that this post will perform better than average, but that to me probably means something like connecting with 7% of the people who read it instead of 5%.

If you don't know that going in and depending on what your expectations are that can be pretty brutal when you realize it (especially if, unlike it sounds like for you, you focus more on the people it doesn't work for that the people it does), and I feel like you did well enough on this post that you might do more and you deserve to know this in case it will affect your self-esteem and your likelihood of doing writing more things like this. Again, this is in the category of "things I wish someone had told me 5 years ago because then I wouldn't have had to figure it out the hard way for myself".

Comment by gworley on Personalized Medicine For Real · 2019-03-05T21:30:05.309Z · score: 6 (3 votes) · LW · GW

No, because there's generally not an option for that via insurance since doing that would effectively be bribery under the way payment is handled. Have not tried doesn't-take-insurance private practice.

Comment by gworley on Personalized Medicine For Real · 2019-03-05T02:38:27.640Z · score: 14 (5 votes) · LW · GW

I have to say I do really wish there were some kind of reliable, N=1 medical service out there for when something is wrong and it's not easy to diagnose let alone solve. I have a lot of personal experience in this area on the patient side, where a person close to me was (and still is!) suffering from some kind of medical problem and they keep getting bounced around because whatever is wrong is rare enough that it doesn't show up on anyone's flowchart. The experience is incredible frustrating, because I can see that there's something pretty specific wrong, but every time I or the patient talked to a doctor we'd go through the diagnostic process and, at best, end with "yep, idk what's wrong, let's just treat some symptoms then". I'd think that we'd be able to do better than this, but in the end most doctors just seem to throw up their hands and say "well, too hard for me, good luck". I get why it happens: it's a lot of work, they're not going to get paid extra for doing it, and no one is going to sue them as long as they made a best effort. But it doesn't make it any less frustrating, and any less interesting (to me) a problem to try to solve both for N=1 and for all the N=1s.

Comment by gworley on Motivation: You Have to Win in the Moment · 2019-03-04T21:18:08.625Z · score: 2 (1 votes) · LW · GW
And if your abstractions are tight (not leaky) enough, you actually don’t need to understand the underlying complexity for them to be useful.

This sounds like the crux of the disagreement: I think no abstraction is sufficiently non-leaky that you don't (eventually) need to understand more of the underlying complexity within the context I see this post sitting in, which is the context of what we might call cognitive, personal, psychological, or personal development (or to put it in non-standard term, the skill of being human). Unless your purpose is only to unlock a little of what you can potentially do as a human and not all of it, every abstraction is eventually a hindrance to progress, even if it is a skillful hinderance during certain phases along the path that helps you progress until it doesn't.

For what it's worth, I also suspect the biggest hurdle we have to overcome to make progress on being better at being humans is gaining enough cognitive capacity to handle more complex, multi-layered abstractions at once, i.e. to see both the machine and the gears at the same time. Put another way, it's gaining the ability to not simply abstract "away" details but to see the details and the abstraction all at once, and then do this again and again with more layers of abstractions and more fine-grained details.

Comment by gworley on Karma-Change Notifications · 2019-03-04T21:02:08.989Z · score: 5 (3 votes) · LW · GW

Oh, thanks! From the post and the comments I thought such a feature didn't exist on purpose!

Comment by gworley on Karma-Change Notifications · 2019-03-03T04:33:45.786Z · score: 3 (4 votes) · LW · GW

I appreciate what you're doing here trying to protect us, but I'd also really like a way to get the data more frequently. I understand there's probably a lot of reasons you want to make this hard, but if it's easy to tweak per user it'd be nice if I could do something like send a support request to get my frequency cranked up to once every 5 minutes or something reasonably real-time that doesn't put a bunch of strain on the system.

Basically, I know I can trust myself with this and would like it, understand why you would want to make it very hard for almost everyone to get access to it, and so just want to put out a feeler to see if super-hidden options are a possibility, even if it means I have to add the code myself and get you to flip the config for my user in the database.

Comment by gworley on Motivation: You Have to Win in the Moment · 2019-03-03T04:24:55.542Z · score: 19 (8 votes) · LW · GW

I think this is the best, mostly clearly and accurately written explanation of this insight to appear within the rationalsphere so far. Most of us, myself definitely included, have focused our explanations largely on narrow ways to approach this point without doing justice to the breadth of it, and I'm not really sure why we've all done that, though my guess is we focus too much on our own entry points, and possibly you've done the same but your way into the insight happened to be one that naturally admits a general explanation. Either way, kudos.

That said, this wouldn't be a very LessWrongy comment if I didn't have few, possibly antithetical, things to say about it.

First, I agree that you get the model right, but it's a model that is only very compelling from a certain stage of development, my strongest evidence being it was once very compelling to me and now it's more like the kind of understanding I would have if I was asked to manifest my understanding without explaining below a certain level of detail, and the other being I think I've seen a similar pattern of discovering this and then focusing on other thing in the writing of others. That doesn't make any of it wrong or not useful, but it does suggest it's rather limited, as I think fellow commenter Romeo also points out. That is, what's going on here is much deeper than it appears to you, and if you keep pushing to explain the opaque parts of this model (like, "where do the beliefs that power motivations come from?" and "why do you even prefer one thing to another?") you'll see it explode apart in a way that will make you go "oh, I had it right, but I didn't really understand it before" the same way you might think you understand how any complex system like a watch or a computer program works until you start literally looking at the gears or electrical currents and then say "oh, I'm amazed I even had such good understanding before given how little I really understood".

I say this not because I want to show off how great I am, even if it seems that way, but because I think you're on the path and want to make it absolutely clear to you that you made progress and that there's much, much deeper to go, whether you pursue that now or later. I say this too because I wish someone had said it to me sooner, as I might have wasted less time being complacent.

Second, just to set expectations, it's unfortunately unlikely that having this model will actually help many people. Yes, it will definitely help some who are ready to see it, but years of trying to explain my insights has taught me that one of the great frustrations is that fundamental insights come in a particular order, they build on each other, and the deeper you go the smaller the audience of people explaining your insights to will help. This doesn't mean we shouldn't do it, as I think anyone who figures these thing out can attest because we've all had both the experience of reading or hearing something of someone else's insight that helped us along and of figuring something out and then helping others see it through our explanations, but it also means we're going to spend a lot of time writing things that people just won't be ready to appreciate yet when they read it. Again, this is a pattern it took me a long while to accept, and once I understood what was going on I overcame much of my previous feelings that I was misunderstanding things despite clear evidence to the contrary because when I tried to explain my understanding it often was met with confusion, misunderstanding, or hostility (my Hegelian writing style not withstanding).

I very much look forward to seeing part 2, and hope it ends up helping many people towards gaining better understanding of how motivations work!

Comment by gworley on Rule Thinkers In, Not Out · 2019-03-01T01:56:44.997Z · score: 5 (3 votes) · LW · GW

You know, at first when I saw this post I was like "ugh, right, lots of people make gross mistakes in this area" but then didn't think much of it, but by coincidence today I was prompted to read something I wrote a while ago, and it seems relevant to this topic. Here's a quote from the article that was on a somewhat different topic (hermeneutics):

One methodology I’ve found especially helpful has been what I, for a long time, thought of as literary criticism but for interpreting what people said as evidence about what they knew about reality. I first started doing this when reading self-help books. Many books in that genre contain plainly incorrect reasoning based on outdated psychology that has either been disproved or replaced by better models (cf. Jeffers, Branden, Carnegie, and even Covey). Despite this, self-help still helps people. To pick on Jeffers, she goes in hard fordaily self-affirmation, but even ignoring concerns with this line of researchraised by the replication crisis, evidence suggests it’s unlikely to help much toward her instrumental goal of habit formation. Yet she makes this error in the service of giving half of the best advice I know: feel the fear and do it anyway. The thesis that she is wrong because her methods are flawed contradicts the antithesis that she is right because her advice helps people, so the synthesis must lie in some perspective that permits her both to be wrong about the how and right about the what simultaneously.
My approach was to read her and other self-help more from the perspective of the author and the expected world-view of their readers than from my own. This lead me to realize that, lacking better information about how the human mind works but wanting to give reasons for the useful patterns they had found, self-help authors often engage in rationalization to fit current science to their conclusions. This doesn’t make their conclusions wrong, but it does hide their true reasoning which is often based more on capta than data and thus phenomenological rather than strictly scientific reasoning. But given that they and we live in an age of scientism, we demand scientific reasons of our thinkers, even if they are poorly founded and later turn out to be wrong, or else reject their conclusions for lack of evidence. Thus the contradiction is sublimated by understanding the fuller context of the writing.
Comment by gworley on Quantifying Human Suffering and "Everyday Suffering" · 2019-03-01T01:47:22.405Z · score: 2 (1 votes) · LW · GW
In the case of humans, it seems self-evident that suffering is a consciously experienced, mental or psychological phenomenon. This makes it difficult to quantify, given our lack of access to other beings’ qualia.

I think we can actually say something about minds in general here. Suffering is tied to how we relate to certain qualia, that is, suffering is qualia about qualia, and to put a fine point on it I'd say suffering is a kind of confusion (an incorrect prediction, in the predictive processing model) we experience as aversive (negative feedback). This suggests that any thing we think shows signs of capacity to perceive its own experience is likely to be capable of suffering, though whether or not it actually suffers is harder to suss out because humans can, though extensive training, learn to not suffer under conditions that would normally cause suffering by changing how they relate to pain.

I realize this doesn't really address the broader point of your post, but since you're thinking about these topics I thought you might find a more precise explanation of suffering of interest since it took me a while to pick t

Comment by gworley on Is LessWrong a "classic style intellectual world"? · 2019-02-26T21:53:53.718Z · score: 13 (5 votes) · LW · GW

I posted this as a question because, although I have my own thoughts on this, I'm not very confident in them because I'm too deep inside LW to be able to see this well, so I want to find out what others think about this. That said, it'd be dishonest if I didn't at least give what I think is the answer, since if nothing else it motivated me to ask the question!

Here's the evidence that jumps to mind when I think about this:

  • certain authors seem to get disproportionately high post scores relative to what I perceive to be the value of what they're saying
    • those authors tend to be authors who I perceive to have high status on LW, often drawn from several sources, such as in person connections, other work they are doing, and previous posts they have published
      • yes, i'm somewhat among this set of people, although it's (thankfully?) often offset by my being so weird that a few people downvote me
      • contra: maybe these people always have something valuable to the LW community to say
    • those authors sometimes write posts they say they think are low quality within the post, yet they still receive higher post scores than average
      • contra: maybe they misassess the quality of their posts
      • contra: maybe they say that for other reasons, like to feel like they've protected themselves from potential rejection if the post is poorly received
  • new authors sometimes write very insightful things that are largely ignored (receive very few votes and have lost post scores)
    • most often i notice this when they fail to conform in some way to LW style expectations
      • contra: maybe writing style is a secretly cherished LW value that isn't held up as explicitly as others like insightfulness and accuracy
    • contra: new people join all the time and some of them gain high status

None of this is perfectly explained by LW being a "classic style intellectual world", but it seems quite suggestive to me that it has tended in that direction, arguably right from the beginning. Maybe the answer will offer other explanations of this evidence that fit it better, or offer other evidence to suggest I have a skewed view of what LW is like (that's why I say I'm too deep inside LW to see it clearly).

Is LessWrong a "classic style intellectual world"?

2019-02-26T21:33:37.736Z · score: 30 (7 votes)
Comment by gworley on Native mental representations that give huge speedups on problems? · 2019-02-26T01:57:30.722Z · score: 7 (4 votes) · LW · GW

A few years ago I wrote a short series of posts on my old blog about what I had learned in this direction. Glancing over it I don't think it's 100% what you're looking for, but might point you in some useful, interesting directions. The posts, in order:

I wrote these when I was at a different stage of cognitive development than I'm in now, so they don't totally match the way I would address these topics today, but hopefully they will be of some use nonetheless.

Akrasia is confusion about what you want

2018-12-28T21:09:20.692Z · score: 18 (15 votes)

What self-help has helped you?

2018-12-20T03:31:52.497Z · score: 34 (11 votes)

Why should EA care about rationality (and vice-versa)?

2018-12-09T22:03:58.158Z · score: 16 (3 votes)

What precisely do we mean by AI alignment?

2018-12-09T02:23:28.809Z · score: 29 (8 votes)

Outline of Metarationality, or much less than you wanted to know about postrationality

2018-10-14T22:08:16.763Z · score: 19 (17 votes)

HLAI 2018 Talks

2018-09-17T18:13:19.421Z · score: 12 (4 votes)

HLAI 2018 Field Report

2018-08-29T00:11:26.106Z · score: 44 (19 votes)

A developmentally-situated approach to teaching normative behavior to AI

2018-08-17T18:44:53.515Z · score: 12 (5 votes)

Robustness to fundamental uncertainty in AGI alignment

2018-07-27T00:41:26.058Z · score: 7 (2 votes)

Solving the AI Race Finalists

2018-07-19T21:04:49.003Z · score: 27 (10 votes)

Look Under the Light Post

2018-07-16T22:19:03.435Z · score: 25 (11 votes)

RFC: Mental phenomena in AGI alignment

2018-07-05T20:52:00.267Z · score: 13 (4 votes)

Aligned AI May Depend on Moral Facts

2018-06-15T01:33:36.364Z · score: 9 (3 votes)

RFC: Meta-ethical uncertainty in AGI alignment

2018-06-08T20:56:26.527Z · score: 18 (5 votes)

The Incoherence of Honesty

2018-06-08T02:28:59.044Z · score: 22 (12 votes)

Safety in Machine Learning

2018-05-29T18:54:26.596Z · score: 17 (4 votes)

Epistemic Circularity

2018-05-23T21:00:51.822Z · score: 5 (1 votes)

RFC: Philosophical Conservatism in AI Alignment Research

2018-05-15T03:29:02.194Z · score: 27 (9 votes)

Thoughts on "AI safety via debate"

2018-05-10T00:44:09.335Z · score: 33 (7 votes)

The Leading and Trailing Edges of Development

2018-04-26T18:02:23.681Z · score: 24 (7 votes)

Suffering and Intractable Pain

2018-04-03T01:05:30.556Z · score: 13 (3 votes)

Evaluating Existing Approaches to AGI Alignment

2018-03-27T19:57:39.207Z · score: 22 (5 votes)

Evaluating Existing Approaches to AGI Alignment

2018-03-27T19:55:57.000Z · score: 0 (0 votes)

Idea: Open Access AI Safety Journal

2018-03-23T18:27:01.166Z · score: 64 (20 votes)

Computational Complexity of P-Zombies

2018-03-21T00:51:31.103Z · score: 3 (4 votes)

Avoiding AI Races Through Self-Regulation

2018-03-12T20:53:45.465Z · score: 6 (3 votes)

How safe "safe" AI development?

2018-02-28T23:21:50.307Z · score: 27 (10 votes)

Self-regulation of safety in AI research

2018-02-25T23:17:44.720Z · score: 33 (10 votes)

The Malicious Use of Artificial Intelligence: Forecasting, Prevention, and Mitigation

2018-02-23T21:42:20.604Z · score: 15 (4 votes)

AI Alignment and Phenomenal Consciousness

2018-02-23T01:21:36.808Z · score: 10 (2 votes)

Formally Stating the AI Alignment Problem

2018-02-19T19:07:14.000Z · score: 0 (0 votes)

Formally Stating the AI Alignment Problem

2018-02-19T19:06:04.086Z · score: 14 (6 votes)

Bayes Rule Applied

2018-02-16T18:30:16.470Z · score: 12 (3 votes)

Introduction to Noematology

2018-02-05T23:28:32.151Z · score: 11 (4 votes)

Form and Feedback in Phenomenology

2018-01-24T19:42:30.556Z · score: 29 (6 votes)

Book Review: Why Buddhism Is True

2018-01-15T20:54:37.431Z · score: 23 (9 votes)

Methods of Phenomenology

2017-12-30T18:42:03.513Z · score: 6 (2 votes)

The World as Phenomena

2017-12-06T02:35:20.681Z · score: 0 (2 votes)

Towards an Axiological Approach to AI Alignment

2017-11-15T02:07:47.607Z · score: 11 (6 votes)

Towards an Axiology Approach to AI Alignement

2017-11-15T02:04:14.000Z · score: 0 (0 votes)

Doxa, Episteme, and Gnosis

2017-10-31T23:23:38.094Z · score: 4 (4 votes)

React and Respond

2017-10-24T01:48:16.405Z · score: 5 (5 votes)

Regress Thyself to the Mean

2017-10-19T22:42:08.925Z · score: 9 (4 votes)

Gnostic Rationality

2017-10-11T21:44:22.144Z · score: 34 (24 votes)

Artificial Unintelligence

2017-10-10T01:37:22.000Z · score: 0 (0 votes)