I strongly endorse reading Impro. It’s short, well-written, and packs a very high insight-to-text ratio.
PeterBorah
Karma: 449
I tried this exercise and found it extremely interesting. My report is rot-13′ed to respect the blinding. I highly recommend taking five minutes to try it yourself before reading.
(One more line break for further encouragement to break your train of thought and try it.)
Ok, here:
V gevrq guvf fvggvat ng zl qrfx ng jbex. V’ir cerivbhfyl unq gebhoyr jvgu rkrepvfrf gung nfxrq zr gb zbir gur ybpngvba bs “zr”/zl rtb/zl crefcrpgvir bhg bs zl urnq naq vagb zl obql, fb V jnf fxrcgvpny V jbhyq fhpprrq. Ohg Dvnbpuh’f qrfpevcgvba va grezf bs “fubbgvat bhg ryrpgevpvgl gb gur erfg bs zl obql gb znxr vg zbir” znqr zr abgvpr n qvfgvapgvba orgjrra gur ybpngvba bs “gur jngpure” be zl bofreingvba/creprcgvba, naq “gur npgbe” be zl zbirzrag/npgvba. Guvf frpbaq guvat ghearq bhg gb or abg gung uneq gb zbir, juvyr gur svefg fgnlrq va cynpr.
V qba’g unir irel fgebat zragny vzntrel, fb V qvqa’g unir nalguvat yvxr n oyhr onyy bs raretl va zvaq. Zbgvba naq fcngvny eryngvbafuvc ner angheny gb zr, fb V sbphfrq ba gur vqrn bs na nofgenpg ybphf gung “chfurq bhg” gb pnhfr zbgvba.
Gb xrrc gur sbphf ba npgvba, V zbirq zl unaq va n ercrgvgvir hc-naq-qbja zbgvba juvyr V gevrq zbivat gur ybphf. Nf fbba nf vg tbg bhg bs zl urnq vagb zl hccre purfg, zl unaq zbirzragf tbg zhpu fybjre, yrff cerpvfr, naq yrff erthyne. Vg sryg n ovg yvxr jnyxvat onpxjneqf, be zbivat jvgu zl rlrf pybfrq. Gurer jnf nyfb n fbeg bs hcjneqf cerffher ba gur ybphf, yvxr jura lbh gel gb chfu fbzrguvat ohblnag qbja vagb jngre.
V ybfg genpx n pbhcyr gvzrf naq unq gb fgneg bire, ohg ba gur guveq gel V zbirq vg vagb zl thg, naq gura gb gur onfr bs zl fcvar. Nf V nccebnpurq gur onfr bs zl fcvar, na ryrpgevp, ohmmvat, fbzrjung birefgvzhyngvat raretl pragrerq gurer orpnzr fgebatre naq zber fnyvrag. Guvf raretl vf snzvyvne gb zr sebz zrqvgngvba, ohg V’ir cerivbhfyl sbhaq vg gbb vagrafr gb “ybbx ng” sbe ybat. Gb zl fhecevfr, gung jnfa’g gehr guvf gvzr, creuncf orpnhfr V jnf pbaarpgvat vg gb zbgvba naq npgvba engure guna whfg ybbxvat ng vg.
Gur raretl nebfr naq frggyrq dhvgr dhvpxyl, va n znggre bs n pbhcyr frpbaqf. Vg jnf (naq vf, nf V jevgr guvf) fgvyy cerfrag naq fnyvrag, ohg irel znantrnoyr. Nf fbba nf vg frggyrq, zl zbgvbaf jrag onpx gb orvat pbasvqrag, snfg, naq cerpvfr. Nsgre whfg n frpbaq, V (jvgubhg pbafpvbhfyl pubbfvat gb) fgbccrq gur ercrgvgvir unaq zbgvbaf naq ghearq gurz vagb n fgergpu, svefg sbe zl unaqf naq nezf, naq gura zl jubyr obql. Zl cbfgher jrag sebz fybhpurq gb fgenvtug-onpxrq (ohg ernfbanoyl erynkrq) naq V fcernq zl yrtf gb n zber pbzsbegnoyr cbfvgvba. V abgvprq gung zl cnagf jrer hapbzsbegnoyl gvtug naq erfgevpgvat, naq nyfb abgvprq gur cyrnfher bs tvivat zl obql n fngvfslvat fgergpu naq ernqwhfgzrag.
Guvf fgngr bs tbbq cbfgher naq obqvyl njnerarff vf snzvyvne gb fbzr qrterr, nf V bppnfvbanyyl trg vagb vg nsgre ubhe-ybat zrqvgngvba fvgf. Vg’f pbaarpgrq gb trareny tebhaqrqarff naq cerfrapr va fbpvny naq zragny ernyzf nf jryy, ohg vg’f n fgngr V unira’g orra noyr gb gevttre irel eryvnoyl.
V tbg hc gb tb gb nabgure ebbz, naq sbhaq zlfrys zbivat zber dhvpxyl naq pbasvqragyl guna V’z hfrq gb. Gurer’f n fjvatvat qbbe orgjrra gjb ebbzf, naq V fbeg bs yrnarq vagb vg jvgubhg oernxvat fgevqr. Zl zbgvbaf erzvaqrq zr n ovg bs ubj V’ir frra nguyrgrf va pnfhny fvghngvbaf zbir, jvgu n fbeg bs pbasvqrapr naq frafr gung gurve obql vf na rssrpgvir gbby. Vg nyfb orpnzr dhvgr boivbhf naq fnyvrag gung V’z uhatel, fvapr V unira’g rngra oernxsnfg.
V’ir xrcg gur ybphf bs npgvba ng gur onfr bs zl fcvar juvyr jevgvat guvf zrffntr. Vg frrzf gb pheeragyl erdhver n fznyy nzbhag bs onpxtebhaq nggragvba gb xrrc vg gurer (fvzvyne gb, yvxr, crepuvat ba n oenapu juvyr ernqvat n obbx, be fbzrguvat yvxr gung), ohg qbrfa’g frrz gb abgnoyl vagresrer jvgu guvaxvat be jevgvat. V srry trarenyyl sbphfrq naq ratntrq, gubhtu vg’f uneq gb gryy jurgure gung’f na rssrpg be whfg n znggre bs orvat vagrerfgrq va guvf rkcrevzrag.
Sure, so now there are two bags:
1) 1000000 white balls and 10 black balls, numbered 1-10.
2) 5 black balls, numbered 1-5.
And now the question is: Bob drew a ball from a bag. Which is more likely?
1) It was a black ball with a number between 1 and 5.
2) It was a black ball with a number between 1 and 10.
~~~
I considered submitting the above as my full response, but here is another approach.
You seem to be substituting a question about the process of choosing for the original question, which was about outcomes. An example where your approach would actually be correct:
“We know that Alice has access to two lists online: an exhaustive list of mathematicians, and an exhaustive list of mathematician-plumbers. We know that Alice invited Bob over for dinner by choosing him from one of those two lists. We know that, by complete coincidence, Alice’s toilet broke while Bob was over. We know that Bob successfully fixed Alice’s toilet. Which list did Alice originally choose Bob from?”
In that case, it’s likely that the Bayesian calculation will say she probably used the Mathematician-Plumber list.
But notice that last question is different from the question of “which of the online lists is Bob most likely to be on?” We know that the answer to that is the Mathematicians list, because he has a 100% chance of being on that list, where he only has a high-probability chance of being on the Mathematician-Plumbers list.
This seems to be a problem of partitioning.
In your analogy, no ball that is in the second bag is also in the first bag. However, all mathematician-plumbers are also mathematicians.
In other words, your analogy is comparing these options:
1) Bob is a mathematician and Bob is not a plumber.
2) Bob is a mathematician and Bob is a plumber.
In that comparison, it is indeed possible that #2 is more likely.
But the actual problem asks you to compare these options:
1) Bob is a mathematician, and Bob either is a plumber or is not a plumber.
2) Bob is a mathematician, and Bob is a plumber.
Since all Bobs in #2 are also in #1, #2 cannot be more likely than #1.
Upvoted for an interesting idea that feels promising. I’d be down to try this experiment, though an hour a day feels like a large time commitment (and regular commitments like that are harder for me to maintain, since my schedule varies wildly).
Proposed alternative:
Once per week you receive an email with a link to a scheduling tool (something like Doodle), where you input your availability for that week. You’re matched with a random partner who has overlapping availability, and you both get an email with the date/time, and a link to a video conference room (perhaps using appear.in) where the call will happen.
I’m uncertain how long the call should be. An hour is not super long, if it’s the only call you’ll have with that person, but people might balk at a longer call. It could be configurable, but also choices are bad.
It would likely be good to have a fairly concrete set of suggestions for what to do on the call. Maybe something like a few minutes of introductions, followed by something like pair debugging? Or if the goal is more about networking, maybe prompts like “share what you’ve been thinking about lately” or “what are your most important goals?” would be good.
I could see value in staying in touch with your partner over email or something for a period of time after the call, but I’m not sure exactly what that should look like, and simplicity is good.
Oh, and an important piece of evidence I forgot to include:
I have a close friend who has a strong “romantic fanfic” habit, and based on observation and conversation it’s extremely clear to me that this is a pica/coping mechanism for them, to imperfectly replace real emotional intimacy/social safety/etc.
The similarity between their behaviors around fanfic and my behaviors around scifi are too strong for me to ignore, so that updates me in the direction of thinking it’s important to uncover what’s going on here.
I would be interested hearing more details about your experiences around exceptions. My inner simulator is confused about how to categorize this particular example.
On the one hand, “no Reddit at home” is a fairly clear rule that I wouldn’t anticipate too much trouble implementing. On the other hand, if the goal is to break the dopamine cycle associated with Reddit, it might be better training for your brain to stop entirely rather than “teasing” it with exceptions?
For the record, even though I’m giving up fiction for Rationalist Lent, I would pretty much agree with that. I intend to use it as an opportunity to break what I consider to be bad habits around it and reevaluate its place in my life, not as a prelude to a permanent ban.
I chose to give up fiction for Rationalist Lent (before reading the comments and before seeing this exchange).
I’ve been observing my fiction (particularly sci-fi) habits for the past year or so, and have tried to reduce/improve them, including making it a major focus both times I attended a CFAR workshop. This has been only marginally successful. The behavior feels addictive in both the sense that I lose sleep or forget other important things due to it, and in the sense that it’s hard to stop.
When I investigate the desire to read scifi, it seems to be connected to desires around intellectual exploration, a sense of wonder/excitement, and some sort of displaced social desire around “camaraderie in service of large-scale goals”. This makes me think it’s a pica, and that I would be better served by using that energy to get more of those things in the real world.
So, all of the above, at least in my case.
This is a good idea.
I hereby commit to giving up reading fiction for the duration of Lent.
It seems to me that you’re conflating “succeeding” and “returning 5000x my money”. I think it’s obviously true that most startups don’t have a 0.1% chance of returning 5000x your money.
Notice that the most important factor in that calculation is generally not “chance of success”, but rather “chance of 5000x return, given success”.
Ok, interesting! I’m updating towards that being a real thing, then.
(It wasn’t obvious from my reading the post that you had explicit evidence for the theory. It came across to me more as “the only thing I can think of”, or something like that.)
I suspect it’s more about trivial inconveniences and failure to “summon sapience”. I was homeschooled and don’t think I’m particularly traumatized about paper, and I love it when I remember to use it, but I have to stop and remember that it’s a good idea, and then go get and use it, which doesn’t always happen.
This is probably not the “simplest” explanation I could give, but I find it memorable. I heard this story in college, but haven’t looked for the reference.
~~~
There was apparently once a study for a new medication, which the researchers really thought would work. So they went through the whole process of a randomized trial, and it came back with no result. The medicine didn’t help.
They thought about this for a while, and then wondered if maybe the problem was that people weren’t taking the medicine regularly enough. Perhaps people were skipping days, or taking it at different times every day, and this was screwing things up? So they re-ran the study, this time tracking compliance. And lo and behold, people who regularly took the medicine did indeed do better than average!
As good scientists, however, they applied the same analysis to their placebo tests. And they found that people who regularly took the placebo also did better than average.
It turns out that the medicine was indeed worthless, but that the sort of people who remember to take their medicine every day are just generally healthier, because they are more likely to be concientious and do things like exercise and eat well.
Yep! When you make 1000-to-1 bets, usually you lose.
a significant chance of making some useful amount of money
It sounds like you’re talking about a different bet than the one the article is about. Gwern’s 0.05% is not “a significant chance”.
I agree that for your totally different case, you should put more effort into security.
[Edit: The 80⁄20 for crypto security is to buy from Coinbase if you’re in the US, or the most credible local exchange if not. If you’re buying altcoins, convert from BTC/ETH on poloniex or shapeshift. Then put it in a hardware wallet such as Trezor or Ledger, with paper copies of your private key in a couple secure locations.]
But if VCs invested in every company with a story they thought was 0.1%-plausible, they wouldn’t make any money.
This is pretty much exactly what they do, if by “0.1%-plausible” you mean “has a 0.1% chance to give us a 10,000x return”. The vast majority of VC investments lose money, and that’s fine, because one 10,000x pays for a lot of failures.
Paul Graham has a good essay on it: http://www.paulgraham.com/swan.html
Now, that probably doesn’t mean VCs shouldn’t filter startups at all (though I’d be interested to see someone try that as an experiment). So it’s worth considering what made Bitcoin “not nonsense”. My first take is that it was a genuine technological innovation, which also got interest from a decent number of smart people. But this is a really unintuitive game. You have to be OK with usually losing all your money. Your system 1 isn’t going to like that.
okay, sure, this one time I might have made a lot of money, but most of the time I would waste a bunch of money/time/other resources.
This seems like a good heuristic, but is it actually true that Bitcoin was an undifferentiated member of an extremely large class? I agree that the availability heuristic is very important here, but I’m struggling to think of a single thing in the same reference class (maybe cryonics? MIRI? Nootropics?). If anyone can give me examples of “stuff lots of people on LessWrong liked that could have huge payouts”, that would be much appreciated.
Security is great! I love security. I recommend hardware wallets if you’re storing a non-trivial amount of crypto.
But the question is about what someone should do when it’s 2011 and they want to buy $10 worth of Bitcoin as a (+EV) lottery ticket. My claim is that, if your goal is “have some Bitcoin”, then the options go like this:
“Buy Bitcoin + implement security best practices” > “Buy Bitcoin without worrying about security” > “Don’t buy Bitcoin”
It’s great if you can get the first one, but it’s irrational to let the existence of the first strategy push you into the third strategy. The second strategy ends up with “maybe some Bitcoin”, which is more Bitcoin then “definitely no Bitcoin”.
Is this a social worry (“people will use it as a blugeon”) or an epistemic worry (“people will incorrectly think there’s a hierarchy, but actually they’re all useful frames”)?
I don’t have strong feelings about shell/shield/staff, but I’ve gotten a lot of value out of Kegan levels, and I think the hierarchy is actually a loadbearing part of the theory. (Specifically, it matters that each level is legible to the one after it, but not vice versa.) I endorse being careful about the social implications, but I wouldn’t want that to become a generalized claim that there aren’t skill hierarchies in the territory.