Discussion [Rationality] An Introduction

https://en.m.wikipedia.org/wiki/Alan_Turing

Please read up on it and stop defaming Sir Alan Turing; he's my best friend's hero. ;(

To cure your ignorance:
https://www.quora.com/What-is-bayesian-rationality
http://lesswrong.com/lw/1to/what_is_bayesianism/

 

1.) Its probably not in your best interest to try and argue with someone that's actually collaborated on research for the MRLI.

2.) Your entire reference is a blog and a book written to introduce common people to complex theories. I've read every book that Hawking wrote, does that make me an astrophysicist now?

3.) You know so little about Bayesian Rationality that you can't even recognize the theory from Alan Turing that was used to develop it. I never said he was associated with Bayesian Rationality, you made that up all on your own.

4.) Philosophy is the science of evidence-based belief formation. Its also the ancestor that all modern specialized sciences evolved from. The part of bayesian rationality theory you're talking about is just a fancy wrapper for "cogito ergo sum", something else you would look into if you ever realized how little you understand about the subject you're pretending to be an expert in.

5.) Your little speech about IQ, I don't know where that came from or why that's relevant. The difference between you and me is that I don't need to run around proving how smart I am to anyone else. My actions reflect that more than enough.

6.) The only real way to learn something is to understand and accept how little you actually know and be willing to learn from any source you can possibly find. I might have actually decided to teach you what I know, if you weren't such a giant douche.

If you ever decide you want to learn something, come find me.
--------------
EDIT: And now he's referencing wikipedia articles at me. All of you that told me to be nicer to him and help him understand, I tried. There's no point in engaging anymore, he's just one of those people that'll never get better.

 

What's this?

What does MRLI stand for?

The book and blog are both written and Authored by Eliezer Yudkowsky. The book is endorsed by the Machine Research Intelligence Institute.

Eliezer Yudkowsky is the poster boy for Bayesian Rationality.

Robin Hanson another CS dude also collaborated with the book.

The book is an accumulation of two years of daily blog posts to www.overcomingbias.com and www.lesswrong.com

Eliezer Yudkowsky is to human rationality, what Stephen Hawking is to Astro Physics.

I think if Stephen Hawking wrote a book on Astrophysics, then it would serve as valid reference no?

At least, I have PROVIDED references. You leave us with nothing but your word. Considering your position on Bayesian Rationality, I find your word doubtful. Your not authoritative enough for me to take what you say for granted.

 

It's more likely I don't know enough about Alan Turing.

I know the basic stuff every CS researcher knows, and some stuff I read on Wikipedia.

From what I know about Epistemological Bayesian Rationality, it was developed using probability theory. I know not of which theory of Turing your referring to.

Please provide REFERENCES.

No. No.

This is not Philosophy. I funnily have little formal training, but Philosophy is a major hobby of mine.

Learn what Philosophy is here: http://www.reddit.com/r/philosophy/wiki/faq

I don't pretend to be an expert. I stated all my knowledge is second hand from Eliezer Yudkowsky. I've stated so right from the very beginning.

You implied you were much much smarter than me. Do you have amnesia?

I'm not trying to prove I'm smart. I was trying to teach something to people. You, we're clearly trying to prove you were smart, and having a little circle jerk going on.

Funny, I'll say the exact same thing to you.

You certainly have things to teach me; about the human brain for example. I find it doubtful you have much to teach me about the subject of this thread. Not because, I'm a connoisseur on the topic, but due to your seeming ignorance and unwillingness to learn about the matter at hand.

Every allegation you levied against me was wrong.

1. I was not a pseudoontellectual.
2. I was not trying to make myself feel smart or convince NUF that I'm smart.
3. I was not another deluded HPMOR reader.
4. I'm actually quite intelligent (MENSA level IQ).

The moment you realised you were mistaken about me, wouldn't it have been prudent to reconsider your position? Redraw your arguments.

You're not even bothering to peruse the links I've provided. I would honestly have ignored you, but it would have affected my potential teaching prospects.

 

I get the impression you've read a single author or source (in this case Ludowsky) and have yet to fully immerse yourself in the full philosophy of reason, wherein a number of dissenting and contradictory views are floated around a lot. This impression has only gotten stronger with your inability to accept contradictory views (@Devshard specifically) and your insistence that there is only one correct answer, which is blatantly untrue in any form of philosophical discourse.

I get that you think Bayesian rationality is absolutely correct and so on, but your dealing with the realm of philosophy. 'Cogito ergo sum' is the only statement in that field which is accepted fact, and even the nature of the cogito and the sum are debated daily. I don't recommend teaching philosophy of any form until you've studied enough that when someone mentions Kant (whose two forms of reason you are usurping), Weber (who is the leader of philosophy of reason), Hume and Descartes (who combined introduced the idea that causality can never be proven) you know exactly what they're talking about, or others like Camus and Neitzsche (nihilists, although I hate both), Freud, Heiddeger, or Wittgenstein you can at least note what their main beliefs are. And, forgive me if I'm wrong, but you don't seem to have any real understanding of any of them.

PS: my list of philosophers is from a non-philosophy perspective. Honestly, there are hundreds if not thousands who should be read and understood, from the ancient Greeks and Chinese to the modern European schools of thought. My base is woefully lacking in substance.

 

You misunderstand me.

@Devshard, specifically attacked me, trying to discreditise me.

@Devshard wasn't floating a contrary opinion.

He was saying "This dude doesn't know what Bayesian Rationality is. This is it".

The validity of Bayesian Rationality as a subject for discussion. However, what it IS, is not.

@Devshard gave a wrong presentation of it. I gave the presentation provided by Eliezer Yudkowsky (The poster boy for the subject).


I fully understand it, don't worry.

 

Ask that in thailand.

 

That's a possibility. One of you sounds like a child angry at his favorite toy being taken away, the other like a scholar who understands conflict and that what is between the lines is more important than the text itself. I don't think I need to point out which is which. That shapes my impression and makes me more likely to believe him than you (also, sources that aren't published are not considered sources in my book. Even ArXiv doesn't always qualify. So linking a blog, even if it is written by someone supposedly at the forefront of a field, does not qualify as evidence to overcome the lack of competent rational argumentation).

So I guess what I'm saying is, even if you do understand what you are trying to present, you don't understand how to present. Which gives the impression that you don't understand what you are trying to present. And when someone, picking up on that the same way I did, puts forth their own understanding in a much more eloquent and thought out format, who is right becomes irrelevant in the face of who seems more capable of discussing the topic at hand.

In the end, my point remains: don't try to teach if you don't understand how to do it, whether from lack of understanding of the subject matter or lack of understanding of how to teach in practice. There are enough bad ideas in the world that there is no need to add more, whether on purpose or not.

 

I know of Immanuel Kant (I have a translation of his book).
I know of David Hume.
I know of Rene Descartes.
I know of Friedrich Nietzsche.
I know of Sigmund Freud.

I know not of Camus.
I know not of Wittgenstein.
I know not of Heiddeger.

I may not be as familar with them as you are, but I do know of them. I do think I'm qualified to teach Bayesian rationality.

I'm also most likely are far more familiar with Yudkowsky's work and his Philosophy than you or @Devshard are.

Which is all that matters in this thread. I have no idea why @Devshard attacked me, but he wasn't challenging the legitimacy of Bayesianism. Nay; he was challenging my personal legitimacy itself.

He also seemingly doesn't know Bayesianism, yet seemingly claims he does, based on what he does know of Scientific inquiry and Rationality.

I do not claim Bayesianism is absolute. Do not misrepresent this situation. I was responding to someone who called me "deluded", "pseduointellectual", "douche" and what else.

I respond more to insults against my intelligence than others; it's my only redeeming feature after all.

Had @Devshard attacked the legitimacy of Bayesian Rationality as you claim he does, I would have been inclined to cordially engage him. However, he did none of that. For some reason, he sought to deligitimise me. I only defended my legitimacy. There was nothing uncordial about my conduct.

 

Now, I was the person being attacked, and I know not to fall to the bias of assuming people see things the same way I do/using myself as a point of reference, yet I can't understand how you came to the conclusion you did.

Even being objective, was it not obvious he sought to merely defame the OP?

Rationality: From AI to Zombies is a book if you're interested. However, it's merely an accumulation of two years of blog posts by Eliezer Yudkowsky and Robin Hanson.

The Book is a compressed compilation of the blog. The blog is more informative.

I understand enough to teach the subject, and I also think I understand how to teach? I'm not too sure.

I'm wondering how you arrived at your conclusion. I'll go reread the thread.

 

@Aca1814ina,

How did you reach your conclusion in light of the above posts? I'm honestly curious.

 

@Aca1814ina - I tip my hat to you sir. There are very few people outside academic philosophy that even know that the "cogito ergo sum" essay is still being debated. Let me know if you ever want to talk about philosophy.

@Dragon God - Sigh. Alright, let's try again.

1.) MLRI standards for "Machine Learning Research Institute". Its what MIRI used to be called a while back, and there's still a couple of papers from them floating around under that name.

2.) Eliezer Yudkowsky did not "invent" Bayesian Rationality out of nowhere, nor is he the representative of the theory. He developed "Bayesian Rationality" theory further based on his own research into artificial intelligence and the fundamental theories Alan Turing developed for his 'Universal Machine'. I get that you're a Comp Sci major, but have you ever read any of Turing's papers or studied his work? Based on the fact that you're quoting a wikipedia entry at me, I'm guessing that you haven't. There's an incredible amount of information outside the shallow descriptions you find on Wikipedia and on blogs.

3.) The definitive work on Bayesian Rationality is this book, written by Mike Oaksford and Nick Chater in 2005 and published in 2007. There are also many articles in academic journals discussing the topic from the 1990's. The blog and book that you keep citing as the authoritative source for all things related to Bayesian Rationality wasn't even conceived when it was being discussed in academic circles.

I actually ran line edits on excerpts from that Oaksford book for another textbook published by a professor at my undergrad university. I can still quote exact passages from there because of how detailed and exact the line edits had to be.

4.) Bayesian Rationality is based almost completely in philosophy. You just don't understand any of the subject matter well enough to know that. Email Eliezer Yudkowsky and ask him to confirm. He'll be the first person to tell the philosophical roots of his work, most of which is based on linguistics. If you'd actually read the blog that you keep throwing around as an absolute defense, you would find several posts discussing the philosophical origins and theories by different philosophers.

5.) We haven't started discussing the validity of Bayesian Rationality, or any real discussion of the topic, because you don't understand it well enough to contribute anything of value to that discussion.

6.) What Bayesian Rationality is, both the classical and Eliezer Yudkowsky's fork-- Its a published theory with evidenced reasoning available for every part of it. No one's debating that. I'm debating your flawed understanding, interpretation, and presentation of it. Almost nothing of what you presented is based on Bayesian Rationality or the many published articles discussing it.

7.) As for MENSA, the only people that throw it around are *guaranteed* pseudo-intellectuals. If you ever actually went to a local or national MENSA meeting or had access to their message boards, you'd understand that its basically just a giant clusterfuck of people that want to talk about how much smarter they are compared to other people. There's no complex problem solving or deep, intense discussion happening there. MENSA articles aren't even eligible for publication in mediocre academic journals, let alone the most prestigious ones.

Don't get me wrong, I was super proud of qualifying when I was 14. Now, I'm pretty embarrassed that I wasted so much money paying that useless organization membership fees and conference fees for five years. Trust me when I say this, MENSA doesn't carry the prestige you think it does. Its more of a badge of shame.

 

You stated Turing killed himself because he gave up on AI research.

That's blatantly false. That's the level of knowledge on Turing you'd expect from someone who needs to read his Wikipedia article.

I haven't read Turing's papers yet no. I'll take a look at them.

I said invent/representative off. Yudkowsky is the poster boy for Bayesian rationality.

He didn't develop it due to AI research(at least, that's not the story from his blog/book).

He's not the originator; he himself holds E. T. Jaynes in high regards as one of the founders of Bayesian Rationality.

Eliezer Yudkowsky is however the poster boy for Bayesian rationality. He's representative of the field. The way Hawkins is of Astrophysics.

 

Maybe it is the originator (and it might not be. E. T. Jaynes work might predate it. I don't know the date of Jayne's work). I've seen the book on Google before.

Nevertheless, Eliezer Yudkowsky is still the poster boy of Bayesian rationality. His work on it, and Lesswrong, is most commonly associated with Bayesian rationality nowadays.

No; I didn't say it wasn't based on Philosophy. I said your definition of Philosophy is WRONG. Blatantly wrong. Even a Highschool introduction wouldn't define Philosophy as such. What you stated, dealt only with Epistemology.

 

I disagree. Well, my understanding of Bayesian rationality is based entirely on Yudkowsky's work, and at best the references he's provided. I don't where your understanding of it comes from.

Your being blatantly inconsistent. First you said I was saying bullshit(and my posts were erroneous. Let's resolve that shall we?) Now your glossing over it, and saying something else. All I said that you said was wrong, was authored by Eliezer Yudkowsky.

So which is it?

Is he a:

1. "smart guy"
2. A liar ?

Let's first deal with all the supposed bullshit I said. After we resolve this, we can continue cordial discussion.

 

Everything I presented is based on Yudkowsky's work.

I actually provided ally references.

So Yudkowsky's representation is flawed?

If that's the case, then this argument is resolved. I'm here to teach Bayesian rationality as presented by Eliezer Yudkowsky.

 

I'm not a MENSA member; there's no branch in my country. I was merely telling you to join them since you have a "one in a billion" g factor.

 

I don't know about Prestige. I do have a friend in the UK who is a MENSA member. he has an IQ of 164, and doesn't tell me anything bad about it.

I was merely saying that I qualify for MENSA.

You may dislike it, I've never been to a meeting, so I don't know. It doesn't seem to be a badge of shame from what little I do know from the net.

 

Because I'm enough of a lady that any curse words are the same to me and I tend to ignore them when others use them. Also, while portions of @Devshard's posts seems childish, the majority does seem to be well reasoned arguments to justify the use of such childish language. Further, that was his opening, which transitioned into the more recent and more rational text. Also a point in his favor is less double-triple-way too many in a row postings. Though you would both be well served to use spoilers.

Further, the debate seems to be about the same point I previously made: Devshard claims that you aren't using a well rounded basis of philosophy but instead the work of one individual ('s blog, specifically), which is automatically incorrect in academic circles (true story, unfortunately. Peer review sucks). You claim you are correctly interpreting that one work with knowledge of the basis yet never appearing to possess that basis. Technically, you're both right in what you claim.

Not a chance. I've made it quite clear that my understanding of philosophy is limited to a tiny portion of metaphysics alone. Epistemology is anything but my strong suit. Also, I'm not a sir.

 

Better, slightly.

Philosophy is the science of reality. I don't use that definition with most people because there are very few people that actually understand what that means. "Cogito ergo sum" is the authoritative statement on what the science of reality is because it concisely explains the Mind-Body-Reality problem.

Epistemology is the study of knowledge and a part of philosophy, and its where all rational discourse and inquiry comes from. Bayesian rationality and its applications are fundamentally an epistemological theory. We make decisions based on our understanding and beliefs at any given point. Bayesian rationality essentially tells us to question that understanding and belief to make better decisions later on.

All of this is still grounded in a deontological and utilitarian framework because you cannot raise any questions into human behavior without using that framework. Instrumental and Value rationality are terms that Max Weber coined around 100 years ago. They deal with an individual's PURPOSE for acting and define relationships based on the end result.

Instrumental Rationality is an outcome-oriented line of reasoning. All interactions with other human beings, living things, objects and the given environment only exist for the purpose of achieving success in a defined goal.

Value Rationality is a line of reasoning in which you're trying to gain something intangible (friendship, conversation, happiness, understanding, etc) through the process of trying to achieve an outcome, and there isn't a significant drive to succeed because what you're looking to gain is on the journey.

Instrumental - success and a focus on the end result.
Value - focus on the process and intangible, success isn't required or something you strive for.

Those are the two concepts you explained poorly. Now let's get down to what they actually mean and why they exist, which is something you completely misunderstood. All of us have both of these 'reasoning circuits' in our brains, and we use them to provide justification for a given purpose. The keyword here is JUSTIFICATION. If you ask yourself why you're doing any given task, its either because it helps you achieve success in a defined end-outcome or because you're gaining something intangible through the process.

Bayesian rationality further extends these theories to create probability statements, or predictions, of your future behavior based on the justification provided for past actions. If you only did certain things in the past using value reasoning as justification and did not really care about achieving the end outcome, then it holds logically true that future actions of a similar nature will also not achieve the end outcome. But you didn't really do those actions for the end-outcome, you already achieved success by gaining an intangible on the process. That's where your little presentation falls apart.