Where do I most obviously still need to say "oops"?

by lukeprog1 min read22nd Nov 201162 comments


Personal Blog

Eliezer once told me:

The most common error I see on Less Wrong is the failure to say "Oops."

If there's one rationality skill I like to think I'm pretty good at, it's this one: the skill of saying "Oops."

In fact, I say "Oops, fixed, thanks" so often on Less Wrong I once suggested I should have a shortcut for it: "OFT."

And I don't just say "oops" for typos and mistakes in tone, but also for mistakes in my facts and arguments.

It's not that I say "oops" every time I'm challenged at length, either. I don't say "oops" until I actually think I was significantly wrong; otherwise, I stand my ground and ask for better counter-arguments.

But I'm sure I can improve.

Wanna help me debug my own mind?

Tell me: On which issues do you think I most obviously still need to say "Oops"?

62 comments, sorted by Highlighting new comments since Today at 6:05 PM
New Comment

Well, since you have asked for feedback, I may provide some, although probably not of the kind requested by this post.

Your repeated requests for feedback accompanied by links proving that you are able to correct your mistakes and like the corrections ... create an impression of some heavy signalling going on. Namely, it's one of the norms here - and probably also among the readers of your blog - to be able to accept constructive criticism, to avoid confirmation bias, to respect a lot of debate rules etc. Signalling adherence to those norms increases one's status. But then, if the signalling part is too apparent, it naturally leads to suspicion of hypocrisy and distrust on the meta-level.

Of course this all is obvious. I write it only because I am not sure whether you realise that the way you ask for feedback may appear to fall into this category. Or actually belong to that category. Substituting the usual pride for never being wrong by pride for not being the sort of person who takes pride from never admitting wrongness is a useful mind hack, but still it's a goal different from being right in the first place.

Not that I could find a single instance where your signalling goals prevented you from finding the truth efficiently. You are certainly not the usual open-mindedness signaller. If there is a danger, it is certainly subtle. Just you don't need to ask for feedback this way. Valuable criticism is usually spontaneous: when people detect a mistake, they say it (unless the local norms discourage such reactions, but that's certainly not the case on LW). On the other hand, when requested to do so, people start to hastily search for something to be criticised, and either find an unimportant detail or even construct a non-existent problem packed in a cloak of rationalisations, or they fail to find anything and produce equally useless you-are-so-awesome-so-that-I-can't-find-a-single-problem response. (Not that it isn't pleasant to hear the latter.)

In short, if you make a mistake, don't be afraid we'll keep it secret.

[-][anonymous]9y 7

I don't actually expect you to get the best feedback by pinging the Discussion board. In addition to reading your content, I've probably interacted with you in person more than most of LW, and I'm still not sure I am qualified to deliver the kind of feedback you'd actually find useful.

It doesn't hurt to ask, of course. But the best feedback I've ever gotten was from people who interact with me daily; people that see me at my worst, see the behavior patterns I'm blind to, and see the inconsistencies in my reactions to different situations and people.

LW will offer laser-focused feedback on your intellectual views, perhaps. But if you want to hear difficult truths that will really help you, it seems like you should turn to those closest to you.

One danger is that you reach out for feedback like this, receive only minor recommendations, and then subconsciously feel as though you've made the motions necessary to be justified in thinking that you're on the right track. I don't expect that this is what you're doing, but I've made simpler mistakes before.

On which issues do you think I most obviously still need to say "Oops"?

I swear I'm not just going meta to look good, but...

This issue, the issue of getting feedback, debugging your mind, improving yourself, your posts, and your style. What you are doing is amazing already; the feedback we can give takes you from the top 99th percentile to the top half of the 99th percentile. But you are very nearly at the point where we can't help you anymore, because we don't know anything that you don't already know and use. Between this and the google survey you ought to get decent data, but beyond these ...

Well, I feel like the biggest difference between you and Eliezer (and I speak of how I perceive you on this site) is apart from subject matter (you obviously can't rehash what he's written and he hasn't the research time and knowledge to do neuroscience) your levels of confidence. Eliezer is self-confident, you ask us for tips to improve. That affects your appearance on the site. So I think you should - not scale down your feedback requests - make your feedback requests stealthier. Maybe in person, in private message, or in comments instead of posts. Maybe focus on collecting metrics data like upvotes or number of dissenting vs approving comments.

Is this something that fits with your observations?

Eliezer is self-confident, you ask us for tips to improve. That affects your appearance on the site.

I agree with this observation, but I'm not sure it's a bad thing. Eliezer is often a bit out of reach to be a proper model: I have often the feeling that I will need countless years to reach his level, I I ever manage to reach it. I can instead relate better to Luke's quest of going from Padawan to Jedi, and I don't mind seeing this kind of behavior from him, at least until he completes his training and becomes a full time member of the Jedi Council.

I feel like the biggest difference between you and Eliezer [is] your levels of confidence.

Yes. Also, our respect for fashion.

I've thought about making my requests for feedback stealthier, but... do I really want to appear more confident than I am? The appearance of overconfidence helps in dating and probably in executive directoring, but there are costs — such as outsiders dismissing me (and SIAI) as falling prey to timelines optimism.

do I really want to appear more confident than I am?

No. You should appear more confident than you currently appear, but you should appear that confident because you are that confident.

"Oops, it looks like half the threads in the discussion section are started by me."

[-][anonymous]9y 1


It is a bad thing if it discourages people you want posting from posting. Which could happen if Luke came off as dominate and territorial. I do not think Luke appears dominate and territorial so this has not registered as a problem to me.

Luke, I don't feel I know you well enough to help you with your quest to locate any lingering wrongness in you. From what I've seen of your writing and what I've heard from people who have met you, you're doing a really amazing job of walking the rationalist talk. The fact that you even ask the community here this question is quite a testament to your taking this stuff seriously and actually using it. I think I should be asking you this question!

But your asking this makes me think of something. If you, or Eliezer, or someone else of that calibre of rational competence pointed out to me an area where I need to say "Oops" (or otherwise direct rational attention), I'd like to think that I'd take that seriously. I suspect I'd take it even more seriously if there were some avenue for me to ask such people for that help the way you've asked the whole Less Wrong community here.

So I wonder: Might it be a good move to set up something like that? We might not yet have a good metric in place for what constitutes someone's degree of rationality, but I'd imagine if two or three black-belt Bayesians all agree that someone is wrong about something, that should still count for something and is probably a reasonable direction to consider in the absence of a more objective metric. So if there were something set up where people could actively ask for that feedback from known people of skilled rationality (or people designated by those with known impressive levels of rationality), I wonder if that would be useful. What do you think? Or would that just be redundant with respect to the Rationality Dojos you mentioned are coming?

If this could be arranged in the future, we'd want to involve top-level non-SIAI rationalists like Julia Galef to avoid results dictated by the SIAI memeplex rather than by rationality skills. (By "top-level" I don't mean "popular" but "seriously skilled in rationality".)

The questions that I really care about, the important questions, are either not solved yet or it is everyone versus the "SIAI memeplex".

All of the following people would disagree with the Singularity Institute on some major issues: Douglas Hofstadter, Steven Pinker, Massimo Pigliucci, Greg Egan, Holden Karnofsky, Robin Hanson, John Baez, Katja Grace, Ben Goertzel...just to name a few. Even your top donor thinks that the Seaststeading Institute deserves more money than the Singularity Institute. And in the end they would either be ignored, downvoted, called stupid or their disagreement a result of motivated cognition.

[This comment is no longer endorsed by its author]Reply

...my downvotes of Ben Goertzel and suggestions that he was stupid are not based on motivated cognition...

Never said that.

The impressiveness of his name (that thing which causes you to refer to him as an authority)...

WTF? I don't think that he has any authority. This essay pretty much shows that he is a dreamer. Two quotes from the essay:

“Of course, this faith placed in me and my team by strangers was flattering. But I felt it was largely justified. We really did have a better idea about how to make computers think. We really did know how to predict the markets using the news.”


“We AI folk were talking so enthusiastically, even the businesspeople in the company were starting to get excited. This AI engine that had been absorbing so much time and money, now it was about to bear fruit and burst forth upon the world!”


That would be awesome, for sure! But I'd also prefer not to see this get frozen in planning just because there's a theoretical possibility of making it better. I'd still consider SIAI-biased advice to be vastly better than no advice at all.

We already have too much lesswrong-exclusive jargon. "OFT" is unnecessary.

Even though if it were accepted it might be OFT used.

That was my decision, too. :)

This isn't really an "oops" but I do think you should spend some time exploring alternative approaches to cognitive science. The usual SIAI position seems to be to act as if there's a single, homogeneous field called "cognitive science" and contrast it with dissenting non-empirical philosophy. At the very least, say, read Tim van Gelder's introductory chapter to "Mind as Motion" to get a sense of the dynamicist critique of computationalism (and, if inclined, look at some of the empirical research) and check out JJ Gibson's "Senses Considered as Perceptual Systems" and "The Ecological Approach to Visual Perception" for ecological psychology. There's also a lot of neuroscience outside cognitive/computationalist neuroscience.

Van Gelder represents computationalism this way:

According to [the computational] approach, when I return a serve in tennis, what happens is roughly as follows. Light from the approaching ball strikes my retina and my brain's visual mechanisms quickly compute what is being seen (a ball) and its direction and rate of approach. This information is fed to a planning system which holds representations of my current goals (win the game, return the serve, etc.) and other background knowledge (court conditions, weaknesses of the other player, etc.). The planning system then infers what I must do: hit the ball deep into my opponent's backhand. This command is issued to the motor system. My arms and legs move as required.

In its most familiar and strike and successful applications, the computational approach makes a series of further assumptions. Representations are static structures of discrete symbols. Cognitive operations are transformations from one static symbol structure to the next. These transformations are discrete, effectively instantaneous, and sequential. The mental computer is broken down into a number of modules responsible for different symbol-processing tasks. A module takes symbolic representations as inputs and computes symbolic representations as outputs.

This is indeed a popular formulation of the computational theory of mind originally defended by Putnam and Fodor, but I'm not sure I've seen it endorsed in so many incorrect details by a major Less Wrong author. For example my post Neuroscience of Human Motivation disagrees with the above description on several points.

I'm not sure the implementation details are particularly relevant to his main argument though. The central concern is that computation is step-wise whereas dynamicism is continuous in time. So a computational approach, by definition, will break a task into a sequence of steps and these have an order but not an inherent time-scale. (It's hard to see how an approach would be computationalist at all if this were not the case.) This has consequences for typical LessWrong theses. For example, speeding up the substrate for a computation has an obvious result: each step will be executed faster. If we have a computation consisting of three steps, S1 -> S2 -> S3, and each one takes 10 ms and we speed it up by a factor of 10 we'll have a computation that executes in 3 ms instead of 30 ms. But if we have a dynamical equation describing the system this isn't the case. I can speak of the system moving between states - say, S(t) -> S(t+1) - but if we speed up the components involved by 10x (say, these are neural states, and we're speeding up the neurons) I don't get the same thing but faster, I get something else entirely. Perhaps the result would be greater sensitivity to shorter time-scales but given that the brain is likely temporally organised I'm inclined to think what I'd get would be a brain that doesn't work at all.

See, this is why I make LW discussion posts asking where I need to say "oops." :)

I encountered dynamical systems when I read my first cogsci textbook, and was probably too influenced by its take on dynamical systems. Here's what Bermudez writes on pages 429-430:

We started out... with the idea that the dynamical systems approach might be a radical alternative to some of the basic assumptions of cognitive science — and in particular to the idea that cognition essentially involves computation and information processing. Some proponents of the dynamical systems approach have certainly made some very strong claims in this direction. Van Gelder, for example, has suggested that the dynamical systems model will in time completely supplant computational models, so that traditional cognitive science will end up looking as quaint (and as fundamentally misconceived) as the computational governor.

[But] as we have seen throughout this book, cognitive science is both interdisciplinary and multi-level... This applies to the dynamical systems hypothesis no less than to anything else. There is no more chance of gaining a complete picture of the mind through dynamical systems theory than there is of gaining a complete account through neurobiology, say, or AI....

...Dynamical systems models are perfectly compatible with information-processing models of cognition. Dynamical systems models operate at a higher level of abstraction. They allow cognitive scientists to abstract away from details of information-processing mechanisms in order to study how systems evolve over time. But even when we have a model of how a cognitive system evolves over time we will still need an account of what makes it possible for the system to evolve in those ways.

Bermudez illustrates his point by saying that dynamical systems theory can do a good job of modeling traffic jams, but this doesn't mean we no longer have to think about internal combustion engines, gasoline, etc.

What do you think?

I think it's essentially begging the question. Van Gelder is questioning whether there is computation going on at all, so to say that dynamical systems abstract away from the details of the information-processing mechanisms is obviously to assume that computation is going on. That might be a way somebody already committed to computationalism could look to incorporate dynamical systems theory but it's not a response to Van Gelder. This is obvious from the traffic analogy. The dynamical account of traffic is obviously an abstraction from what actually happens (internal combustion engines, gasoline, etc). But the analogy only holds with cognitive science if you assume what actually happens in cognitive systems to be computation. What Van Gelder is doing is criticising computationalism for not be able to properly account for things that are critical to cognition (such as evolution in time). It's not clear to me what it could mean to abstract away from computational models in order to study how systems evolve over time if those models do not themselves say anything about how they evolve over time. I think Van Gelder addresses this. It's difficult to get an algorithmic model to be time-sensitive.

That said, whether the dynamical approach alone is adequate to capture everything about cognition is another matter. There are alternative approaches that provide an adequate description of mechanisms but that are more sensitive to the issue of time. For example, see Anthony Chemero's Radical Embodied Cognitive Science where he argues that we need ecological psychology to make sense of the mechanisms behind the dynamics. Typically dynamicists operate on a embodied/ecological perspective and don't simply claim that the equations are the whole explanation (they are concerned with, say, neurons, bodies, the environment, etc). I think Bermudez is also confused about levels here. Presumably the mechanism level for cognition is the brain and its neurons, and perhaps the body and parts of the environment, and a computational account is an abstraction from those mechanisms just as much as a dynamical equation is. It's common in computationalism to confuse and conflate identifying the brain as a computer with merely claiming that a computational approach gives an adequate descriptive account of some process is the brain. So, for example, I could argue that an algorithm gives an adequate description of a given brain process because it is not time sensitive and can therefore be described as a sequence of successive states without reference to its evolution in time. But that would not imply that the underlying mechanisms are computational, only that a computational description gives an adequate account.

But that would not imply that the underlying mechanisms are computational, only that a computational description gives an adequate account

Could you elaborate what you mean by this? Our most successful computational models of various cognitive systems at different levels of organization do remarkably well at predicting brain phenomena, to the point where we can simulate increasingly large cortical structures.

I read most of Van Gelder's last article on dynamical cognitive systems before he switched to critical thinking and argument mapping research, in BBS, and I'm still not seeing why computationalism the and dynamical systems approach are incompatible. For example, Van Gelder says that a distinguishing feature of dynamical systems is it quantitative approach to states - but of course computationalism is often quantitative about states, too. Trappenberg must be confused, too, since his textbook on computational neuroscience talks several times about dynamical systems and doesn't seem to be aware that they are somehow at odds with his program of computationalism.

Naively, it looks to me like the dynamical systems approach was largely a rection to early versions of the physical symbol system hypothesis and neural networks, but if you understand computationalism in the modern sense (which often includes models of time, quantitative state information, etc.) while still describing the system in terms of information processing, then there doesn't seem to be much conflict between the two.

Even Chemero agrees:

On our view, dynamical and [computational] explanation of the same complex system get at different but related features of said system described at different levels of abstraction and with different questions in mind. We see no a priori reason to claim that either kind of explanation is more fundamental than the other.

Thanks for the specific recommendations.

Put CSA out its misery?

Turn it into somewhere to post a link and synopsis to your LessWrong posts.

The quality of commenting there is just really awful compared to the comments on your LessWrong posts, but it has a separate audience that LW doesn't. Even a muddle-headed audience that reads the thing may produce useful results. (They might come here.)

It's possible that I should just start deleting all comments that are shitty, but that would require lots of time.

Only stuff you'd delete as abusive rubbish. But the sincere stupidity is sincere. Best treat it like comments on a newspaper site ;-) You clearly have a readership there (aggrieved theists who think they're philosophers), and surely there's something you can do with that.

[-][anonymous]9y 2


I don't think that is a good use of your time.

Establishing a hierarchy of moderators would be a good idea. Those who become moderators will become invested in the site and its ideas.

I don't want to even invest that much effort in CSA. It's really just a place for me to post occasional articles that don't fit elsewhere, and to post weekly links.

Are you seeing it as being in "misery" because you are expecting it to be what it once was instead of the much more modest, personal, and erratic thing I have declared that it now is?

I am wondering if you have good reasons for investing any time there at all. It's possible. That's why I phrased it as a question.

There are still a surprising number of readers.

Is that your true rejection?

I think so. If it wasn't so instrumentally useful due to its continuing readership, I'd just freeze the whole site and let it sit there as an archive and never have to delete spam comments again.

So you think you can reach a wider audience because many of those people will (for whatever reason) not follow you here.

I think another reason could be: Because one day I want to write a book.

I'm reminded of my recent brainstorming session on that neutrino business.

I'm not sure why such errors wouldn't be addressed in their normal contexts right where they've been made.

[-][anonymous]9y 0