Discussion [Rationality] An Introduction

@Devils Advocate
you get me

 

Of course I "get" you. We are practically clones. It took me a while, but I think Dragon god definitely got it after you did your "lol" and pointed it out that it was a funny thing to say.

 

I personally doubt that Yudkowsky would produce something better on the subject of human rationality than RAZ, but that's not for me to determine.

Must all magnum opus references be used post humously though? Can I not call it his magnum opus while he has yet to write better?

 

Hmmm... can you? I also want to know... run it through your philosophy, it is a 'Rational' thing to claim?

 

Wtf are you implying your first sentence makes no sense at all or it is purely personal beliefs that comes from experience and the same applies to the rest of what you're saying !
Either you explicitly tell your axioms or what you said makes no sense at all to a vast majority.
Also rationality isn't subjective at all only the axioms are (or if you're doing a deductive fallacy)

 

It's more convention, than an issue with rationality.

Are Magnum Opus only assigned post humously

 

Nope I didn't get. I thought he was making a joke about the length of the work.

That's evidence for another bias: Expecting short inferential distances. Assuming that because you "get it", or "it's obvious" to you, that others will "get it" or it'll be "obvious" to them.

 

It's a runaround statement I never had a point or basis to begin with.

 

Omg this post literally shat on my brain xp

 

My dear sir, rationality is just simple; "if it acts like a girl... and looks like a girl... it is a girl!" I don't know if that isn't subjective.

 

I actually skipped your post. Now I feel guilty, and will go back and read it up.

Epistemological rationality should be objective.

Instrumental rationality may be subjective as people have different values.

 

Objective overuled!

I don't know any of those but true, Instrumental rationality is subjective because it is a matter of values. Epistemological, I'll look it up a bit.

 

*shrug*
I think it's up to you I guess, words have meaning and they are what they are.

So sure you can say, NO I don't think the author will EVER come up with something better, then go head call it his magnum opus. Makes you a narrow-minded douche bag, but no one will stop you.

You can say, no I don't mean magnum opus, then don't use magnum opus, cause you don't know what magnum opus means.

Don't say apple when you mean oranges, but apple means apple, not anything else

I don't care that much about any of this BTW, I just like talking to @Casi

 

And here I thought this Magus-whatevs are a self-concept created by someone deeply accepted by people of the preceding generation.

 

I do know what Magnum Opus means. I'm not being a narrow minded douche bag. There are several reasons why it's unlikely Yudkowsky would author a better work on human Rationality. It's unlikely he would write another book on human rationality to even start with.

So, I very well feel he wouldn't author one without being a "narrow minded douche bag".

I also fully know what Magnum Opus means. I'm using it in the correct usage of the term.

 

@Dragon God
look, like I said, I don't care about any of this, you don't have to stress how much you know what Magnum Opus means by saying it twice.

And just so you know, there are millions of narrow-minded douche bag in the world whether you are one or are not one is not a big deal. If it is important to you to not be known as a narrow-minded douche bag, just don't do narrow-minded douche bag things, and people will not think you are a narrow-minded douche bag,

But honestly, I don't see why one does not want to be a narrow-minded douche bag, especially most narrow-minded douche bags; will never realise they are narrow-minded douche bags; because they are narrow-minded douche bags.

I mean if I am a narrow-minded douche bag, I would not know I am a narrow-minded douche bag. Until someone pointed it out.. well since I am entertaining the possibility of being a narrow-minded douche bag, it probably means I am not one.

 

Well, it seemed you were using the term derogatorily. I don't take well to perceived insults.

 

Sigh. Alright. Since my good friend @Dragon God asked me so nicely to prove that he's a pseudo-intellectual that literally has no idea what he's talking about, I'm going to do just that. To make this more interesting, I'm going to do this by teaching all of you what he's hopelessly trying to muddle through. As a special treat for Dragon God, I'm going to do this only using rational discourse. So we're going to start with a basic foundation and build our way up.

Let's start with Bayes Theorem, which is a set of probability measures and tools. The purpose of these tools is to quantify the probability of an event occurring. We can then use these quantified probability measures to establish a relationship between two independent events, and estimate the probability of them occurring together or in a series. Let's use Dragon God's example of flipping a coin to illustrate this. There are two events that can occur when you flip a coin: it can come up heads or tails. So we call the probability of heads appearing P(A) and the probability of tails P(B). Since there are only two possible outcomes, both P(A) and P(B) are equal to 50. Now, if we wanted to know if P(A) would affect the future likelihood of P(B) occurring, we can use some of Bayes Theorem to test that hypothesis. Obviously, we know that no relationship exists between P(A) and P(B) in the coin-flip example.

Now, we'll move onto the scientific method. Some of you might be confused at this point and believe that Bayes Theorem and the scientific method aren't linked. That is false. Bayes Theorem was developed as an extension to the scientific method, and is commonly used in all academic fields for that purpose. Allow me to illustrate with a simple example. Let's say I develop a brand new drug that cures a disease. There's obviously going to be some side effects to taking this medication, which is something that patients need to be aware of before they start using it. In the drug testing process, I start with a total pool of 50,000 people. Only some of them develop specific side effects. We'll say that a common side effect for this drug is diarrhea, and 12,000 out of the total 50,000 people had diarrhea after using this drug. I use the patient history of these 12,000 people to discover commonalities and create a probability statement for a future patient using this drug that might develop diarrhea. Something like this:

P(Diarrhea) = P(Patient age > 35) + P(vegetarian diet) + P(using anti-hypertensive drugs) + P(past history of gallstones)
I can also apply a multiplier to each of the individual contributing events based on how much they contribute to the likelihood of the event Diarrhea occurring.

In essence, Bayes Theorem is used to establish a cause-and-effect relationship between a few independent events and a single outcome.

Now that we're all clear on the basics, let's move onto this whole "Bayesian rationality" stuff. You have to understand, the scientific method and Bayes theorem only apply in isolated systems under a specific set of conditions. This bayesian rationality stuff is an attempt to apply those fundamental concepts to human behavior, which does not exist in an isolated system nor does it have specific conditions. Human behavior has way too many variable and possible outcomes, so it isn't possible to apply the scientific method or bayes theorem to predict how a person might act in any given situation. So the guy that Dragon God keeps going on about came up with a *really* smart way of doing it based on Calculus.

Fundamentally, calculus is a set of tools that allow you to determine the properties of a curve using the same rules that apply to straight lines. I know what you're thinking: a curve and a straight line are two very different things, so how is this possible? Calculus allows us to break a curve down into tiny straight lines. Then we determine the properties of all of those straight lines and put them together. Derivatives are the break down process and integration is putting all the properties we discovered back together to create a single statement for the curve.

"Bayesian rationality" does the same thing for human behavior. You observe how a person behaves in an isolated set of situations under specific conditions. Based on that, we can determine the probability pattern for their overall behavior. Once we've got that, you can determine the probability for how a person will act in any given situation at any given time and figure out what the most likely outcome/choice they'll make is. We can also do this for ourselves, and hopefully make better decisions. Mind you, none of this stuff is even close to being exact. We know next to nothing about how the human brain functions, let alone all the intricacies of human behavior.

A lot of the theory behind "Bayesian rationality" was actually developed by Alan Turing in his efforts to create the universal machine, or what we now call artificial intelligence. The brain works like a giant database, with billions of connections between points of data. Let me demonstrate how these data connections evolve from childhood to adulthood. We learn about different animals as children, like cows. The points of data for cow are:

1.) Four legs
2.) White and black spots
3.) Udders
4.) Goes "moooo"

Those four points are linked to "Cow", which is also linked to "Farm" and all its associated data points, which is also linked to "Milk" and all its associated data points like color-texture-taste-bottling-purchasing, which is also linked to "store" and its associated data points, which is also linked to "beef" and its associated data points, which is also linked to "sustainable farm practices" and all its data points, which is also linked to the "meat industry" and all its associated data points and so on.

I'm only talking about linking concepts right now. I haven't even gotten to linking associated memories and emotions yet. The human mind is this giant web of billions of interconnecting data points that modify each other based on the order of recall, and we're constantly refining and changing the data connection pathways.

When Alan Turing finally started to understand the sheer complexity and scale of the human mind, he gave up his attempt to create true artificial intelligence and committed suicide. Bayesian rationality involves calculating through all of the possible data connection pathways and figuring out the most probable ones to determine the most probable set of behaviors to determine the most probable outcomes over a large series of actions.

The point here is that 85% of all human beings are simply not smart enough to even comprehend the scale of mental calculations needed to attempt using bayesian rationality to predict behavior. 99% of human beings are not smart enough to do it on the spot. It seems completely plausible when you see Sherlock Holmes do it on TV or in the books, but that's only because everything is scripted. Our brains just haven't evolved to the point where its possible.

The other, more interesting, point is that our brains actually do pull this off to a certain extent with people that we know well. Its an entirely subconscious process that occurs between the Amygdala, Cortex, and areas 41 and 44 on the left and right sides of the brain. There's also some involvement from the visual cortex and the pre-frontal cortex, but I'd need several days to explain that bit. That's how we know the effect our speech and actions will have the people close to us, and part of our self-censoring mechanisms.

I'm distilling very complex concepts to keep things simple so everyone understands, but I think you've all got the idea now. Now children, let's bring this all full-circle to a giant moral ideal: Never posture or spew pseudo-intellectual babble at someone much, much smarter than yourself.

-This concludes this week's episode of 'Devshard Debunks Stupidity'. Tune in next week for an all new episode!-

 

On the off chance that someone's interested, Areas 40, 41 and 44 of the brain are also known as Broca's and Wernicke's areas. The two neural circuits are responsible for speech, comprehension, and writing. But that's only the simple part of what they do.

Language, or linguistics really, are a huge part of why humans are so different from other animals. We use language to encode and transmit all of the complex data connections that only exist in our individual minds. When you explain something to someone else using language, it forces you to examine all of your own connections at the neural tissue level and discard useless connections. We also use language to refine our neural connections and understanding of abstract concepts.

Essentially, we can only do things like building cities and create technology because we can talk to each other and build collective knowledge. Language is the fundamental building block of human society.

 

This should be 50% or 0.5. Just saying 50 is wrong.

Never said the scientific method doesn't use Bayesian theorem. The distinction I made between Bayes theorem and Traditional rationality was in the eligibility of the Hypotheses being tested.

Bayesian rationality isn't only about human behaviour. What we're mainly concerned with for now is the epistemological aspect of Bayesian rationality, which deals with forming correct, consistent, accurate beliefs.

Probability theory is used in belief forming.

Through the application of Bayes theorem and known information.

You fucked up.

Lying about Alan Turing in front of a CS major and someone interested in AI.

Turing was a CS researcher yes. He's not associated with Bayesian rationality though.

Turing committed suicide due to pressure and unease from his prosecution due to being homosexual.

I do not know if your smarter than me(I consistently test in the third SD), but that is irrelevant to the point.

Having an IQ of 180, does not mean you know about Bayesian rationality.

Intelligence != knowledge.

Nothing I've said is "pseudointellctual babble".

Unlike you, I actually provided references to sites where you can check everything I've written.

But you sir must be a rare individual. I mean your "much much smarter" than me, and I'm in the 3rd SD. Considering that you should be at least 2 SD above me, you may very well be in the 6th SD. Amazing; you're rarity is on the same order of magnitude as 1 in a billion.

Why don't you put your AMAZING 6th SD intelligence to good use and do something more productive than embarrass yourself over something your not even bothered to research? Join Mensa, Triple nine, Sigma, Giga, etc.

You actually seem quite knowledgeable; about the human brain at least. Your killing yourself with your abject ignorance about the core subject matter though.

You failed to prove this. Better luck next time.