The Dilbert Challenge: you are working in a company in the world of Dilbert. Your pointy-haired boss comes to you with the following demand:

"One year from today, our most important customer will deliver us a request for a high-quality reliable software system. Your job and the fate of the company depends on being able to develop and deploy that software system within two weeks of receipt of the specifications. Unfortunately we don't currently know any of the requirements. Get started now."

I submit that this preposterous demand is really a deep intellectual challenge, the basic form of which arises in many different endeavors. For example, it's reasonable to believe that at some point in the future, humanity will face an existential threat. Given that we will not know the exact nature of that threat until it's almost upon us, how can we prepare for it today?

On the Care and Feeding of Rationalist Hardware

Many words have been spent here in improving rationalist software -- training patterns of thought which will help us to achieve truth, and reliably reach our goals.

Assuming we can still remember so far back, Eliezer once wrote:

But if you have a brain, with cortical and subcortical areas in the appropriate places, you might be able to learn to use it properly. If you're a fast learner, you might learn faster - but the art of rationality isn't about that; it's about training brain machinery we all have in common

Rationality does not require big impressive brains any more than the martial arts require big bulging muscles. Nonetheless, I think it would be rare indeed to see a master of the martial arts willfully neglecting the care of his body. Martial artists of the wisest schools strive to improve their bodies. They jog, or lift weights. They probably do not smoke, or eat unhealthily. They take care of their hardware so that the things they do will be as easy as possible.

So, what hacks exist which enable us to improve and secure the condition of our mental hardware? Some important areas that come to mind are:

• sleep
• diet
• practice

Incommensurate thoughts: People with different life-experiences are literally incapable of understanding each other, because they compress information differently.

Analogy: Take some problem domain in which each data point is a 500-dimensional vector. Take a big set of 500D vectors and apply PCA to them to get a new reduced space of 25 dimensions. Store all data in the 25D space, and operate on it in that space.

Two programs exposed to different sets of 500D vectors, which differ in a biased way, will construct different basic vectors during PCA, and so will reduce all vectors in the future into a different 25D space.

In just this way, two people with life experiences that differ in a biased way (due to eg socioeconomic status, country of birth, culture) will construct different underlying compression schemes. You can give them each a text with the same words in it, but the representations that each constructs internally are incommensurate; they exist in different spaces, which introduce different errors. When they reason on their compressed data, they will reach different conclusions, even if they are using the same reasoning algorithms and are executing them flawlessly. Futhermore, it would be very hard for them to discover this, since the compression scheme is unconscious. They would be more likely to believe that the other person is lying, nefarious, or stupid.

There is a topic I have in mind that could potentially require writing a rather large amount, and I don't want to do that unless there is some interest, rather than suddenly dumping a massive essay on LW without any prior context. The topic is control theory (the engineering discipline, not anything else those words might suggest). Living organisms are, I say (following Bill Powers, who I've mentioned before) built of control systems, and any study of people that does not take that into account is unlikely to progress very far. Among the things I might write about are these:

• Purposes and intentions are the set-points of control systems. This is not a metaphor or an analogy.

• Perceptions do not determine actions; instead, actions determine perceptions. (If that seems either unexceptionable or obscure, try substituting "stimulus" for "perception" and "response" for "action".)

• Control systems do not, in general, work by predicting what action will produce the intended perception. They need not make any predictions at all, nor contain any model of their environment. They require neither utility measures, nor Bayesian or any other form of in

We are Eliza: A whole lot of what we think is reasoned debate is pattern-matching on other people's sentences, without ever parsing them.

I wrote a bit about this in 1998.

But I'm not as enthused about this topic as I was then, because then I believed that parsing a sentence was reasonable. Now I believe that humans don't parse sentences even when reading carefully. The bird the cat the dog chased chased flew. Any linguist today would tell you that's a perfectly fine English sentence. It isn't. And if people don't parse grammatic structures to just 2 levels of recursion, I doubt recursion, and generative grammars, are involved at all.

I'm kind of thinking of doing a series of posts gently spelling out step by step the arguments for Bayesian decision theory. Part of this is for myself: I've read a while back Omohundro's vulnerability argument, but felt there were missing bits that I had to personally fill in, assumptions I had to sit and think on before I could really say "yes, obviously that has to be true". Some things that I think I can generalize a bit or restate a bit, etc.

So as much as for myself, to organize and clear that up, as for others, I want to do a short series of "How not to be stupid (given unbounded computational power)" In which in each each post I focus on one or a small number of related rules/principles of Bayesian Decision theory and epistemic probabilities, and gently derive those from the "don't be stupid" principle. (Again, based on Omohundro's vulnerability arguments and the usual dutch book arguments for Bayesian stuff, but stretched out and filled in with the details that I personally felt the need to work out, that I felt were missing.)

And I want to do it as a series, rather than a single blob post so I can step by step focus on a small chunk of the problem and make it easier to reference related rules and so on.

Would this be of any use to anyone here though? (maybe a good sequence for beginners, to show one reason why Bayes and Decision Theory is the Right Way?) Or would it be more clutter than anything else?

This doesn't even have an ending, but since I'm just emptying out the drafts folder

Memetic Parasitism

I heard a rather infuriating commercial on the radio today. There's no need for me to recount it directly -- we've all heard the type. The narrator spoke of the joy a woman feels in her husband's proposal, of how long she'll remember its particulars, and then, for no apparent reason, transitioned from this to a discussion of shiny rocks, and where we might think of purchasing them.

I hardly think I need to belabor the point, but there is no natural connection between shiny rocks and promises of monogamy. There was not even any particularly strong empirical connection between the two until about a hundred years ago, when some men who made their fortunes selling shiny rocks decided to program us to believe there was.

What we see here is what I shall call memetic parasitism. We carry certain ideas, certain concepts, certain memes to which we attach high emotional valence. In this case, that meme is romantic love, expressed through monogamy. An external agent contrives to derive some benefit by attaching itself to that meme.

Now, it is important to note when describing a Dark pattern t... (read more)

Buddhism.

What it gets wrong. Supernatural stuff - rebirth, karma in the magic sense, prayer. Thinking Buddha's cosmology was ever meant as anything more than an illustrative fable. Renunciation. Equating positive and negative emotions with grasping. Equating the mind with the chatty mind.

What it gets right. Meditation. Karma as consequences. There is no self, consciousness is a brain subsystem, emphasis on the "sub" (Cf. Drescher's "Cartesian Camcorder" and psychology's "system two"). The chatty mind is full of crap and a huge waste of time, unless used correctly. Correct usage includes noticing mostly-subconscious thought loops (Cf. cognitive behavioral therapy). A lot of everyday unreason does stem from grasping, which roughly equates to "magical thinking" or the idea that non-acknowledgment of reality can change it. This includes various vices and dark emotions, including the ones that screw up attempted rationality.

What rationalists should do. Meditate. Notice themselves thinking. Recognize grasping as a mechanism. Look for useful stuff in Buddhism.

Why I can't post. Not enough of an expert. Not able to meditate myself yet.

Aumann agreements are pure fiction; they have no real-world applications. The main problem isn't that no one is a pure Bayesian. There are 3 bigger problems:

• The Bayesians have to divide the world up into symbols in exactly the same way. Since humans (and any intelligent entity that isn't a lookup table) compress information based on their experience, this can't be contemplated until the day when we derive more of our mind's sensory experience from others than from ourselves.
• Bayesian inference is slow; pure Bayesians would likely be outcompeted by groups that used faster, less-precise reasoning methods, which are not guaranteed to reach agreement. It is unlikely that this limitation can ever be overcome.
• In the name of efficiency, different reasoners would be highly orthogonal, having different knowledge, different knowledge compression schemes and concepts, etc.; reducing the chances of reaching agreement. (In other words: If two reasoners always agree, you can eliminate one of them.)

This would probably have to wait until May.

Willpower building as a fundamental art. And some of the less obvious pit falls. Including the dangers of akrasia circumvention techniques which simply shunt willpower from one place to another and overstraining damaging your willpower reserves.

I need to hunt back down some of the cognitive science research on this before I feel comfortable posting it.

Some bad ideas on the theme "living to win":

• Murder is okay. There are consequences, but it's a valid move nonetheless.
• Was is fun. In fact, it's some of the best fun you can have as long as you don't get disabled or killed permanently.
• Being a cult leader is a winning move.
• Learn and practice the so called dark arts!

What would a distinctively rationalist style of government look like? Cf. Dune's Bene Gesserit government by jury, what if a quorum of rationalists reaching Aumann Agreement could make a binding decision?

What mechanisms could be put in place to stop politics being a mind-killer?

Why not posted: undeveloped idea, and I don't know the math.

I'm vaguely considering doing a post about skeptics. It seems to me they might embody a species of pseudo-rationality, like Objectivists and Spock. (Though it occurs to me that if we define "S-rationality" as "being free from the belief distortions caused by emotion", then "S-rationality" is both worthwhile and something that Spock genuinely possesses.) If their supposed critical thinking skills allow them to disbelieve in some bad ideas like ghosts, Gods, homeopathy, UFOs, and Bigfoot, but also in some good ideas like cryonic... (read more)

Yet another post from me about theism?

This time, pushing for a more clearly articulated position. Yes, I realize that I am not endearing myself by continuing this line of debate. However, I have good reasons for pursuing.

• I really like LW and the idea of a place where objective, unbiased truth is The Way. Since I idealistically believe in Aumann’s Agreement theorem, I think that we are only a small number of debates away from agreement.

• To the extent to which LW aligns itself with a particular point of view, it must be able to defend that view. I don’t w

A criticism of practices on LW that are attractive now but which will hinder "the way" to truth in the future; that lead to a religious idolatry of ideas (a common fate of many "in-groups") rather than objective detachment. For example,

(1) linking to ideas in original posts without summarizing the main ideas in your own words and how they apply to the specific context -- as this creates short-cuts in the brain of the reader, if not in the writer

(2) Use of analogies without formally defining the ideas behind them leads to content not o... (read more)

Winning Interpersonally

cousin_it would like to know how rationality has actually helped us win. However, in his article, he completely gives up on rationality in one major area, admitting that "interpersonal relationships are out."

Alex strenuously disagrees, asking "why are interpersonal relationships out? I think rationality can help a great deal here."

(And, of course, I suppose everone knows my little sob-story by now.)

I'd like to get a read from the community on this question.

Is rationality useless -- or worse, a liability when deal... (read more)

I'd be interested in reading (but not writing) a post about rationalist relationships, specifically the interplay of manipulation, honesty and respect.

Seems more like a group chat than a post, but let's see what you all think.

The ideal title for my future post would be this:

How I Confronted Akrasia and Won.

It would be an account of my dealing with akrasia, which so far resulted in eliminating two decade-long addictions and finally being able to act according to my current best judgment. I also hope to describe a practical result of using these techniques (I specified a target in advance and I'm currently working towards it.)

Not posted because:

1. The techniques are not yet tested even on myself. They worked perfectly for about a couple of months, but I wasn't under any severe str

Regarding all the articles we've had about the effectiveness of reason:

Learning about different systems of ethics may be useless. It takes a lot of time to learn all the forms of utilitarianism and their problems, and all the different ethical theories. And all that people do is look until they find one that lets them do what they wanted to do all along.

IF you're designing an AI, then it would be a good thing to do. Or if you've already achieved professional and financial success, and got your personal life in order (whether that's having a wife, having... (read more)

Any interest in a top-level post about rationality/poker inter-applications?

The Implications of Saunt Lora's Assertion for Rationalists.

For those who are unfamiliar, Saunt Lora's Assertion comes from the novel Anathem, and expresses the view that there are no genuinely new ideas; every idea has already been thought of.

A lot of purportedly new ideas can be seen as, at best, a slightly new spin on an old idea. The parallels between, Leibniz's views on the nature of possibility and Arnauld's objection, and David Lewis's views on the nature of possibility and Kripke's objection are but one striking example. If there is anything to... (read more)

Scott Peck, author of "The Road Less Travelled", which was extremely popular ten years ago, theorised that people became more mature, and could get stuck on a lower level of maturity. From memory, the stages were:

1. Selfish, unprincipled
2. Rule- following
3. Rational
4. Mystical.

Christians could be either rule-following, a stage of maturity most people could leave behind in their teens, needing a big friendly policeman in the sky to tell them what to do- or Mystical.

Mystical people had a better understanding of the World because they did not expect it ... (read more)

I have this idea in my mind that my value function differs significantly from that of Elizier. In particular I cannot agree to blowing up Huygens in that Baby-Eater Scenario presented.

To summarize shortly: He gives a scenario which includes the following problem:

Some species A in the universe has as a core value the creation of unspeakable pain in their newborn. Some species B has as core value removal of all pain from the universe. And there is humanity.

In particular there are (besides others) two possible actions: (1): Enable B to kill off all of A, with... (read more)

Putting together a rationalist toolset. Including all the methods one needs to know, but also, and very much so the real world knowledge that helps to get along or ahead in life. Doesnt have to be reinvented, but pointed out and evaluated.

In short: I expect members of the rationality movement to dress well when its needed. To be in reasonable shape. To /not/smoke. Know about positive psychology. Know about how to deal with people. And find ways to be a rational & happy person.

memetic engineering

The art of manipulating the media, especially news, and public opinion. Sometimes known as "spin-doctoring" I guess, but I think the memetic paradigm is probably a more useful one to attack it from.

I'd love to understand that better than I do. Understanding it properly would certainly help with evangelism.

I fear that very few people really do grok it though, certainly I wouldn't be capable of writing much relevant about it yet.

I have more to say about my cool ethics course on weird forms of utilitarianism, but unlike with Two-Tier Rationalism, I'm uncertain of how germane the rest of these forms are to rationalism.

I have a lot to say about the Reflection Principle but I'm still in the process of hammering out my ideas regarding why it is terrible and no one should endorse it.