In that case, I’m an ENTJ.
Nornagest
Karma: 10,894
But without medicalizing, how can we generate significant-sounding labels for every aspect of our personalities?
There’s always divination. It’s totally random, of course, but throw enough parameters and different methods at the problem and eventually most people will hit something they’re happy with.
I’m a Cancer with Aries rising. What’s your sign?
Welcome to LW! I suspect you’ll find a lot of company here, at least as regards thinking in unwarranted detail about fictional magic systems.
I suspect most rationalists will turn out to care more about their cuddle piles than about their ideas becoming mainstream. There’s always been a rather unhealthy interaction between community goals and the community’s social quirks (we want to raise the sanity waterline → we are saner → our quirks should be evangelized), and we don’t really have a working way to sort out what actually comes with increased rationality and what’s just a founder effect.
I think the cost is higher than you’re giving it credit for. Securing dev time to implement changes around here is incredibly hard, at least if you aren’t named Eliezer, and changes anywhere are usually harder to back out than they are to put in; we can safely assume that any change we manage to push through will last for months, and forever is probably more likely.
An earlier version of my comment read “LW or parts of it”. Edited it out for stylistic reasons and because I assumed the application to smaller domains would be clear enough in context. Guess I was wrong.
Granted, not everything I said would apply to the first proposal, the one where top-level posts are upvote-only but comments aren’t. That’s a little more interesting; I’m still leery of it but I haven’t fully worked out the incentives.
As to empirics, one thing we’re not short on is empirical data from other forums. We’re not so exceptional that the lessons learned from them can’t be expected to apply.
Proposals for making LW upvote-only emerge every few months, most recently during the retributive downvoting fiasco. I said then, and I continue to believe now, that it’s a terrible idea.
JMIV is right to say in the ancestor that subtle features of moderation mechanics have outsized effects on community culture; I even agree with him that Eliezer voiced an unrealistically rosy view of the downvote in “Well-Kept Gardens”. But upvote-only systems have their own pitfalls, and quite severe ones. The reasons behind them are somewhat complex, but boil down to bad incentives.
Imagine posting as a game scored in utility. Upvotes gain you utility; downvotes lose you it; and for most people being downvoted costs you more than being upvoted gains you, though the exact ratio varies from person to person. You want to maximize your utility, and you have a finite amount of time to spend on it. If you spend that time researching new content to post, your output is low but it’s very rarely downvoted. Debate takes a moderate amount of time; votes on debate are less reliable, especially if you’re arguing for something like neoreaction or radical feminism or your own crackpot views on time and dimension, but you’re all but guaranteed upvotes from people that agree with you. Plus telling people they’re wrong is fun, so you get some bonus utility. Finally, you can post cat pictures, which takes almost no time, will score a few upvotes from people that like looking at their little jellybean toes, but violates content norms.
Which one of these is optimal changes, depending on how tolerant you are of downvoting and how good you are at dodging it. But while removing the downvote option incentivizes all three (which is why social media likes it), it should be clear that it incentivizes the last two much more. You can see the fruits of this on Facebook groups, that site’s closest analogy to what’s being proposed here. (Tumblr, and Facebook user pages, are also upvote-only in practice, but their sharing and friending mechanisms make them harder to analyze in these terms.)
Prison gangs formed from a kind of arms race of mutual self-defense.
Lots of gangs form that way—it’s one of the two main pathways to organized crime, the other one being economies of scale in selling illegal goods and services. The Bloods, for example, started out as a sort of anti-Crips self-defense force, and many yakuza organizations are generally thought to have their roots in mutual-protection societies among small commercial enterprises.
You’ve cited some examples of people who, it is undeniable, are successful, but who also happen to fit your argument. But equally there are many successful businesspeople who did not study maths/CS/physics
If only there were some way of quantifying this.
I’m not sure if performance-enhancing drug use (at least at the professional level) is a good example of irrationally short planning horizons. I’ll bet Lance Armstrong regrets getting caught, but I’ll also bet he’d have been worse off in the long run (financially, and also fame-wise) if he didn’t use the stuff: it’s not unlikely to have made the difference between “world-famous cyclist in disgrace over a drug scandal” and “peaked at a #5 finish in the 1999 Tour”.
Sure, he lost medals, but that could also be phrased as “bragging rights”, and his cycling career was already over the hill at the time.
Diversity of products is nice to have, but it’s not the main reason economics says a competitive marketplace works better. Instead, it’s important mainly because it provides a dynamic incentive for allocating goods efficiently and against artificially creating rents (in the economic, not the housing, sense). Incentivizing innovation is also important, though, and that could plausibly be linked to the financial crisis: the mortgage-backed securities that played a key role were quite novel. Not that I think incentivizing innovation goes in the loss column overall.
“Diversifies risk” is just bizarre, at least outside a few specific segments: an agricultural region takes on less risk if it grows several different kinds of crops, because then the sector as a whole is less vulnerable to e.g. disease or infestation, but that’s just as true if there’s one farmer running the place or fifty, and it’s not like there’s a species of aphid that feeds on refrigerators.
But what do you expect from a link that includes the phrase “human sheep”?
A finite number of thoughts implies an end to subjective experience. Zeno’s paradox works because distances in the thought experiment can be infinitely subdivided.
Who said anything about a few years? If you paid attention in high school, the linear algebra background you need is at most a few months’ worth of work. I was providing a single counterexample, not saying that the full prerequisite list (which, if memory serves, is most of a CS curriculum for your average ML class) is always necessary.
I can’t imagine a serious situation (as opposed to e.g. a programming contest) where I would have to write my own sort routine from scratch
You can’t? I’ve had to do that several times. The usual scenario is that there are search/sort routines, but they have inconvenient properties—either they don’t perform well in the specific problem domain I’m dealing with (happens a lot in simulation; functions for efficiently doing certain types of categorization on spatially arranged data are rare outside graphics libraries), or they don’t work on the data types I need and a reduction is impractical for one reason or another, or they exist but can’t be used for legal reasons. Unless you always situate yourself in the most popular subfields, which I frankly find boring, you can’t always count on there being a library that does exactly what you want—all the more so in a still-emerging space like ML.
(I’ve never had to build a washing machine, incidentally, but I’ve had to fix washing machines—twice this year for two different machines, in fact. I could have hired a mechanic or bought a new machine, but either one would have cost me hundreds of dollars.)
I wasn’t pointing strictly to research, but I was pointing to low-level implementation. It now occurs to me that I might be unusual in this respect—much of my ML experience is in the context of a rather weird environment that didn’t have any existing libraries, leaving me to cut a lot of code myself.
So I might have to back off from “ability to do machine learning”. You can, in retrospect, use ML perfectly competently in a lot of settings even if the closest you’ve ever gotten to a simulated annealing algorithm is plugging the cost function into a Python library; but I have a hard time calling someone an expert if they’ve never written anything lower-level, just as I’d expect a competent software engineer to be able to write a hash table by hand even if every environment they’re likely to encounter will have built-in implementations or at least efficient libraries for it.
On the other hand, if you don’t have a solid grasp of linear algebra, your ability to do most types of machine learning is seriously impaired. You can learn techniques like e.g. matrix inversions as needed to implement the algorithms you’re learning, but if you don’t understand how those techniques work in their original context, they become very hard to debug or optimize. Similarly for e.g. cryptography and basic information theory.
That’s probably more the exception than the rule, though; I sense that the point of most prerequisites in a traditional science curriculum is less to provide skills to build on and more to build habits of rigorous thinking.
Interesting. But there’s a couple of features in there that make me leery of relying on this.
First, the table there is tracking convictions at jury. Clearance rates, especially for property crime, have never to my knowledge been high; if we assume similar figures they’d be undercounting reported crimes (most comparable to the crime statistics we’re familiar with) by a factor of three to five, and undercount committed crimes by more. That isn’t necessarily a good assumption, though, and there’s the rub: we can’t use conviction rates to estimate crime rates unless we know something about how likely cases were to make it through the system.
Second, buried in the bottom of that page there’s a sentence saying that about 17,000 summary convictions (excluding some minor fines) were imposed independently by police magistrates. No information on type, which means the table would further undercount crimes by some unknown proportion depending on how likely summary punishment was. But 17,000 is roughly four times the total jury convictions cited, so it’d probably be large.
I haven’t been able to find another source going back to the 1830s yet, but this data suggests that as many murders were recorded in London in the late Victorian era as in the mid-Sixties, when population was about 20% higher. (Population in 1838 was much lower—the city grew hugely over the 19th century.)
Not that AD&D does any better, but if you’re in a fight and you’ve exchanged blows twenty times without serious damage, at least one of the fighters isn’t trying to win. Realistic personal combat (other types of combat can be more drawn out) is like Hobbes’ state of nature, or a cannibal elf: nasty, brutal, and short.
The problem is—well, there are several problems here, but the main problem is that it’s really hard to build a multiplayer game that’s actually fun but that looks and plays like realistic fighting; the winners haven’t had time to play, and the losers feel betrayed. You can do it in single-player games, where there’s an endless supply of mooks to shank and no particular requirement that killing your avatar take you out of the fight for more than a few seconds, but you usually end up with a very hard game.
Aiming for cinematics rather than realism is indeed the correct approach, but this wasn’t clear in the tabletop RPG world for a long time; I suspect that has something to do with its roots in the strategy wargame genre, which values realism very highly and has the structure to mostly get away with it.
Theoretically? No reasons other than those I’ve given above. But empirically it’s awfully rare to do that successfully, and people have tried. About the only ones I’ve been happy with are written by our own Scott Alexander (of slatestarcodex), and his politics are close enough to mine that I don’t totally trust my judgment there; I’m more libertarian than he is, but that’s a relatively petty difference given what he tends to write about.
The more common outcome is that a writer sets out to build with steel but inadvertently builds with straw, not so much maliciously as through honest misunderstanding of the opposition, and is only confirmed by the exercise in their existing beliefs. We’re very good at fooling ourselves into thinking that more or less subtle caricatures accurately represent our opponents’ motives.
And I expect I’ll probably catch some flak for saying so, but I don’t have much faith in LW’s ability to move past that stage.
When I was in high school, I was a skinny nerd that could barely bench-press the bar. But I spent most of my senior year eating my lunches with some guys from the football and track teams, including a lineman who went on to the NFL.
These guys weren’t dumb. They generally weren’t academic stars—they did well enough in school not to embarrass themselves in college applications, but they spent their time on the field instead of studying and it showed in grades. But they were quick and clever and could enjoy an intelligent conversation—often a more intelligent one than the nerdy clique, once you’d exhausted the possibilities of Warhammer and Counterstrike.
(A couple years later, I discovered fencing.)