4 ms·
FWIW, it seems to me that a major benefit of the Bayesian approach is to make bad reasoning (in the form of, say, an unreasonable prior) transparent and obvious
by XFrequentist 11y ago
FWIW, it seems to me that a major benefit of the Bayesian approach is to make bad reasoning (in the form of, say, an unreasonable prior) transparent and obvious. I've never heard it claimed that the Bayesian approach was robust to sophisticated idiocy (neither on LessWrong nor mainstream writing on Bayesian methods), except in the narrow techical sense that the posterior asymptotically approximates the likelihood given infinite data (provably true under some assumptions but irrelevant to the objections in the OP).
- TeMPOraL 11y agoYup. Actually, a common theme on LessWrong was realizing that with better reasoning tools you're more able to bullshit yourself, and so you need to be extra-careful.
- zepto 11y agoHow do you be extra careful except by developing yet more powerful reasoning tools?
- TeMPOraL 11y agoYou can't protect yourself in 100% - it would require developing more powerful reasoning tools in an infinite regression. But what you can do is to use introspection, and triple-check your reasoning when it seems to defy common sense or leads you to weird (awful) conclusions. That's why LW is so big on biases and heuristics by the way - you can treat them as a list of warning signs; if your reasoning seems to match some of them, it's time to take a closer look. (A thing a lot of people missed since knowing logical fallacies started to become a mainstream thing - you should use the list to check your own reasoning, not your opponent's.)
- Chathamization 11y agoThe problem with trying to rely on heuristics to avoid biases is people often ignore the biases in the heuristics of choice. To continue the example of LW, there are many people there who seem to think highly of IQ test, and who ignore the many issues with them (the Flynn effect an the effect of incentives being a couple examples of the flaws in IQ tests). Trying to remove biases is great. But there is a problem when someone works to remove some biases, then believes that they are inherently more rational than the public at large, and then uncritically accepts there other biases ("Someone like me who's worked hard to remove there biases must be correct when compared to the biased masses.").
- bordercases 11y agoYou should probably know what James Flynn thinks of the Flynn effect. He doesn't try to escape the conclusions of an I.Q. test as much as extend them, despite some of the "paradoxes".
- Chathamization 11y agoFair enough. It's also worth pointing out that Alfred Binet, generally considered to be one of the fathers (if not THE father) of intelligence testing, felt that the idea of quantitative intelligence testing was severely flawed (and had fairly harsh words for people who believed that there was largely a single measure of intelligence).
- TeMPOraL 11y agoYes, there is that risk, and no doubt many fall for it. Ego / self-esteem issues may be a big part of it. But then again, every worthy goal poses risks. When you fly a plane, there's a greater risk you'll kill yourself than when you stay on the ground, and yet airplanes are being flown and we're reaping great benefits from it. RE IQ, personally, I'm 100% confused on the topic. I used to believe that Flynn effect is basically people getting better at doing tests, but recently I heard that someone controlled for that and the effect remained. So I don't know. The topic is complicated and most of studies I heard of are the kind of psychology and social science I implicitly assume is mostly bullshit.
- danieltillett 11y agoThe Flynn effect mystery is exactly the same as the mystery that in most countries the younger population is taller than the previous generation. The fact that the Flynn effect only occurs in the bottom half of the intelligence distrubtion is a bit of a clue as to what is the cause.
- sawwit 11y agoWhat's the cause that follows from that it mostly affects the bottom 50%?
- cossatot 11y agoSpend more time with the data, literature, alternate hypotheses, and so forth.
- Estragon 11y agoYou don't take anything hammered out by reason alone as sound information. You test reasonable-looking propositions against experience, and until they've stood up against that test by providing accurate predictions with substantial information-content, you take them as provisional.
- NN88 11y ago"The first principle is that you must not fool yourself and you are the easiest person to fool." - Richard P. Feynman
- irotatori 11y agoA fascinating illustration of this being the phenomenon of Roko's basilisk.
- 1024core 11y ago> I've never heard it claimed that the Bayesian approach was robust to sophisticated idiocy Alas, nothing is robust to sophisticated idiocy.
- 1812Overture 11y ago"If you make something idiot proof, someone will just make a better idiot."
- XFrequentist 11y agoThis made me chuckle.
- Estragon 11y ago"The trouble with fool-proof methods is, you're still a fool."
- Retric 11y agoBayesian seems great when you first see it. It should be obvious how to apply it for something like a card game. The problem is how long would it take you to realize a deck of cards was missing the 4 of diamonds? What if the card was lost 1/2 way though the game? How about on the prior hand? In the end it's stuck at one level of recursion and all facts are fuzzy.
- stuartaxelowen 11y agoIt's stuck in the reality you define - you can add the last hand, all cards seen so far, and the color of people's jackets to the model if you so please.
- Retric 11y agoYour still stuck with: You observed X/Y. How accurate is your count. How accurate is your estimate of accuracy. How accurate is your estimate of accuracy of your accuracy estimate. ... recursive infinity.
- cobaltblue 11y agoDid you account for the possibility you're dreaming, or in a simulation? Or in a simulation of a simulation and so on? Yes, that's the curse of real numbers. In practice you set a limit of precision required, and a threshold of evidence you need to tell you that the precision should be more precise. You can get quite far with Newtonian physics without considering the corrections of general relativity, you can get quite far in engineering by assuming linearity by only expanding the Taylor transform of a non-linear expression by a bit instead of infinitely, and you can do a lot by assuming probabilistic independence of your card counting from what's going on on Pluto right now.
- monkmartinez 11y agoIt is going to take me more than a few minutes to parse this. However, I am relatively certain there is a lot of sarcasm in this snippet and I like it.