This is fantastic. Well researched, fairly well written.
I have a niggling general complaint about how LW seems to use rationality as just a general good word. It just, icks me a bit. I suspect that it might really turn off new readers.
Seriously, my one complaint is that when reading this on an iPad it took me too long to scroll past all the references.
I agree that it would be nice for articles with long bibliographies to have a show/hide options (starting hidden). I am unsure how this would be possible at the moment, so a “skip to end” link might have to do.
What would you prefer? Instrumental Rationality is a bit of a mouthful, Common Sense is an abused term is means whatever the speaker believes in, and our super dictionary “Acting as to maximize expected utility” seems formal. I agree we pepper the word Rationality enough that it may turn off outsiders, but I am personally not seeing other terms or phrases that don’t either under formalize to the point of meaningless folksiness or over formalize to the point of turning even more people off.
I found the article quite interesting as a new reader, for what it’s worth. Would love to see more in this vein as well as the more formal or abstract articles about rationality, bias, etc.
I actually don’t like the word “rationality” at all considering that it is most commonly used to mean rationalization which is of course not rationality. If somebody is “rational” I usually think of it to mean “he has justified his actions to himself” or “he has common sense” (common sense being a terrible thing). I prefer the word “logic” as mathematician, but maybe that’s associated with proofs instead of probabilities or something.
Well, the two have a different meaning to me. Logic is a class of mathematical systems, like first-order logic. But logic stays within an axiomatically constructed system, it doesn’t claim or pretend to have a direct link with “reality”.
Rationality is the art of using such a system in relation to reality—to understand reality, predict it, and therefore gain the power to steer it in a preferred direction.
Logic itself will never tell you if the universe uses Newtonian or relativistic laws of motion. Both systems are logically consistent. But rationality will tell you that relativistic laws of motion are a closer map of reality than Newtonian laws of motion (but that Newtonian laws of motion is still a very valid map for daily life).
Yea, this is a good explanation. Logic seems to be considered as more abstract rules, while rationality seems to apply it to reality.
Although considering Bayes Theorem is a logical, mathematical construct, I could certainly argue against the idea that “logic doesn’t claim or pretend to have a direct link with reality”.
Well, I would say Bayes Theorem itself is purely logical, but realizing how it applies to updating your belief network and scoring your hypothesis, and then using it that way, especially the part of devising tests that could falsify your hypothesis, is rationality. I knew Bayes Theorem before discovering Less Wrong, I even knew a bit about Bayesian networks in computer science, but I never realized how deep Bayes Theorem was (and how it was a more powerful, more technical version of the scientific method) before reading the “intuitive explanation” and the Sequences.
But of course, the two are far from totally isolated. Words are fuzzy boundaries, not precise definitions.
This is fantastic. Well researched, fairly well written.
I have a niggling general complaint about how LW seems to use rationality as just a general good word. It just, icks me a bit. I suspect that it might really turn off new readers.
Seriously, my one complaint is that when reading this on an iPad it took me too long to scroll past all the references.
I can’t wait to read more of this.
Yeah, there should be a “skip to comments” link before the bibliography, or a show/hide button or something.
I agree that it would be nice for articles with long bibliographies to have a show/hide options (starting hidden). I am unsure how this would be possible at the moment, so a “skip to end” link might have to do.
What would you prefer? Instrumental Rationality is a bit of a mouthful, Common Sense is an abused term is means whatever the speaker believes in, and our super dictionary “Acting as to maximize expected utility” seems formal. I agree we pepper the word Rationality enough that it may turn off outsiders, but I am personally not seeing other terms or phrases that don’t either under formalize to the point of meaningless folksiness or over formalize to the point of turning even more people off.
If you find that “Rational” belongs at the beginning of most posts, then it can go entirely unsaid.
Much like as I realized just recently, we really don’t need a symbol for “such that” in
∃x(Px)
This.
I found the article quite interesting as a new reader, for what it’s worth. Would love to see more in this vein as well as the more formal or abstract articles about rationality, bias, etc.
I actually don’t like the word “rationality” at all considering that it is most commonly used to mean rationalization which is of course not rationality. If somebody is “rational” I usually think of it to mean “he has justified his actions to himself” or “he has common sense” (common sense being a terrible thing). I prefer the word “logic” as mathematician, but maybe that’s associated with proofs instead of probabilities or something.
But whatever, just sematics and definitions.
Well, the two have a different meaning to me. Logic is a class of mathematical systems, like first-order logic. But logic stays within an axiomatically constructed system, it doesn’t claim or pretend to have a direct link with “reality”.
Rationality is the art of using such a system in relation to reality—to understand reality, predict it, and therefore gain the power to steer it in a preferred direction.
Logic itself will never tell you if the universe uses Newtonian or relativistic laws of motion. Both systems are logically consistent. But rationality will tell you that relativistic laws of motion are a closer map of reality than Newtonian laws of motion (but that Newtonian laws of motion is still a very valid map for daily life).
Yea, this is a good explanation. Logic seems to be considered as more abstract rules, while rationality seems to apply it to reality.
Although considering Bayes Theorem is a logical, mathematical construct, I could certainly argue against the idea that “logic doesn’t claim or pretend to have a direct link with reality”.
Well, I would say Bayes Theorem itself is purely logical, but realizing how it applies to updating your belief network and scoring your hypothesis, and then using it that way, especially the part of devising tests that could falsify your hypothesis, is rationality. I knew Bayes Theorem before discovering Less Wrong, I even knew a bit about Bayesian networks in computer science, but I never realized how deep Bayes Theorem was (and how it was a more powerful, more technical version of the scientific method) before reading the “intuitive explanation” and the Sequences.
But of course, the two are far from totally isolated. Words are fuzzy boundaries, not precise definitions.
It’s only ‘deep’ if you have to dredge it up and out from a pile of bullshit. ;)
Applying Bayes Theorem is just applying logic to your life. It follows directly from the theorem. That would make you logical.
Or perhaps we are just differentiating from the abstract and the real.
How about “decision theory”?