7 ms·
The game theory of John von Neumann transformed the 20th century
- melling 4y agoThis book has been out for several months and has been submitted many times on HN. Perhaps some of these reviews are also interesting? https://hn.algolia.com/?dateRange=pastYear&page=0&prefix=false&query=The%20Man%20From%20the%20Future&sort=byPopularity&type=story https://hn.algolia.com/?dateRange=pastYear&page=0&prefix=fal...
- bee_rider 4y agoMy favorite bit of trivia about von Neumann is that, in between basically going around an inventing the scaffolding for like half of modern science and technology he invented… Mergesort? What a mundane little algorithm to be invented by such a genius. It is the best sorting algorithm though. It is the most elegant, no weird edge cases, and stable.
- 082349872349872 4y agoMergesort with punched card decks would have been fairly obvious, so I wouldn't be surprised if the general idea predated his software implementation by at least several decades?
- bee_rider 4y agoI’m not sure I see the connection to punched card decks, other than that they are a thing that might need to be sorted… but then, people have been sorting things by hand for all of history. He and his colleague are credited with the design of the algorithm… suppose it is possible they drew general inspiration from how people might sort something by hand.
- SamReidHughes 4y agoPunched cards were generally processed by machines in a streaming fashion.
- wolfram74 4y agoBut if you dropped them, you'd need to put them back in order, and decks are very amenable to splitting and dealing.
- marcosdumay 4y agoIt's a very numerous thing that needs to be sorted often, and composed of items with similar shapes. That combination isn't very common. But I wouldn't be surprised if somebody invented it 2 or 3 millennia ago either.
- bee_rider 4y agoI think they were actually talking about the data access pattern, with mergeSort you get nice sequential access to the arrays of course. I don’t actually know how data was fed into card based systems, I mean I know they fed the cards in, but I assumed the bulk data was read from tapes or something and cards mostly held the programs. Maybe not though, it is all quite a bit before my time!
- MichaelMoser123 4y agocould a tabulating machine do recursion? Did it have a stack?
- eru 4y agoYou can implement a bottom-up merge sort iteratively fairly easily. It's basically two nested loops only.
- halpmeh 4y agoIsn’t radix sort way better for punched cards (and more obvious)?
- eru 4y agoDepends what your sort key is. Radix sort works only for small-ish integers, but not well eg for names that you want to sort alphabetically. Nor for floating point numbers or even large integers.
- bee_rider 4y agoCut the cards to correspond to their magnitude and you can just use spaghetti sort.
- veltas 4y agoWell which is it? Mundane or elegant? To me, it being the most elegant makes it a worthy algorithm.
- bibanez 4y agoIndeed. I interpreted OC's choice of words as "mundane" meaning "quotidian" though
- bee_rider 4y agoI don’t think those are incompatible. If Michelangelo designed a paperweight, it would probably be a magnificent paperweight, while still just being a paperweight. (This is an extreme example, Mergesort is better than a paperweight).
- corvus-cornix 4y agoThey never mention the economist he collaborated with in inventing game theory: Oskar Morgenstern. There were also collaborators involved in developing "Von Neumann" architecture. He was just the biggest name involved. He was a genius, but there's some politics at work in who is remembered for collaborative intellectual work.
- chestervonwinch 4y ago> there's some politics at work in who is remembered politics, or do people have natural a bias towards giving attribution to a lone genius?
- pixelpoet 4y agovon Neumann wasn't a lone genius, he was a god, the only "human" completely beyond the bounds of our comprehension. Accepting that is the first required step to understanding his contributions to the modern world.
- dEnigma 4y agoIf anything you're losing some understanding when you try to see him as something other than human. Absolutely outstanding, sure, but not something else entirely.
- pixelpoet 4y agoDon't take it from me: > Nobel Laureate Hans Bethe could not comprehend von Neumann's incredible intellect: "I have sometimes wondered whether a brain like von Neumann's does not indicate a species superior to that of man." He's certainly not alone in that opinion either, cf. Oppenheimer and the rest of "the martians" [1]; I've read several books about him, and the history of mathematics in general. Of course we may disagree, but... [1] https://en.wikipedia.org/wiki/The_Martians_(scientists) https://en.wikipedia.org/wiki/The_Martians_(scientists)
- 4y ago
- kristianp 4y agoThe subtitle of this article "How the game theory of John von Neumann transformed the 20th century." isn't addressed in this article unless briefly in passing. Poor editing (if that's the word for it). This part is intriguing: > In his Theory of Self-Reproducing Automata (published posthumously in 1966), von Neumann went even further, describing the conditions under which, with no more than eight parts (four structural and four dedicated to logical operations), complex creatures could emerge capable of executing any type of computation and even of replicating themselves. Von Neumann machines are a fun part of science fiction. The phrase "Hegemonic swarm" from Iain Banks comes to mind.
- joe_the_user 4y agoYeah, I think that's because the subheading basically isn't true. Game theory is more of footnote on von Neuman's many amazing achievements than a capstone. It influenced cold war thinking certainly, but the situation of the US and USSR didn't really fit the zero sum, repeated games that Von Neuman originally formulated studied. Essentially, the US and USSR faced a "Mexican standoff", "mutually assured destruction"(MAD), a non-repeated, non-zero-sum situation you can claim has a mathematical solution but which ultimately is more of a psychological problem.
- eru 4y agoIf you want to model MAD via game theory, it would certainly be a repeated game. They faced each other day in day out year after year. I agree that war, both cold and hot, is a non-zero sum game. That's why things like temporary ceasefires can work at all. Or Geneva conventions.
- voldacar 4y agoIt is a serious indictment of our culture that we are not trying to clone von Neumann or even just sequence his genome. His genetic material is slowly decaying in some cemetery in New Jersey, I think.
- worik 4y agoGood. RIP
- lordnacho 4y agoDon't you think there's some significant non genetic component to whatever made him so smart? Otherwise we'd be hearing about his relatives.
- doetoe 4y agoMaybe his intellect was due to a mutation that he didn't pass on...
- chimprich 4y agoThere's usually some regression to the genetic mean for children from people who are extreme outliers. Even if you have "smart" genes, you're unlikely to luck out on such an outstanding arrangement of genetic material again... and von Neumann was a very extreme outlier. By some accounts, the most intelligent human in history [0]. You'd expect for any of his kids to be outstanding, but not earth-shatteringly so. And looks like that's about right if you look up his daughter [1]. [0] Check out his "known for" section in Wikipedia! https://en.wikipedia.org/wiki/John_von_Neumann https://en.wikipedia.org/wiki/John_von_Neumann [1] https://en.wikipedia.org/wiki/Marina_von_Neumann_Whitman https://en.wikipedia.org/wiki/Marina_von_Neumann_Whitman
- OkayPhysicist 4y agoIt doesn't help that the aristocratic educational system of having a handful of tutor/mentors as the primary means of education has fallen (steeply) out of fashion. People greatly underestimate just how effective children can learn. They're basically capable of absorbing as much information as you can pump into them. It's a delicate balance making sure they learn enough basic human operating stuff in addition to everything else, but the results can be remarkable. I remember hanging out at camp with a kid who loved board games. Especially probability based ones, such liar's dice and Dominion, but also more general hidden information games like Stratego. Kid was undeniably a prodigy, and he kicked my ass every time. I had just graduated university with studying CS and Physics, this kid was looking forward to starting kindergarten in the fall.
- photochemsyn 4y agoI for one am tired of the endless hagiography surrounding JvN, who while certainly a brilliant mathematician was also a certifiable wannabe architect of genocide on a massive scale. This was fairly well known a few decades ago, for example (1992): https://www.americanheritage.com/prisoners-dilemma https://www.americanheritage.com/prisoners-dilemma > "In 1950 Life magazine quoted the great Hungarian-American physicist John Von Neumann, co-father of both the atom bomb and the digital computer, advocating immediate pre-emptive nuclear war against Russia: “If you say why not bomb them tomorrow, I say why not today? If you say today at 5 o’clock, I say why not one o’clock?” He was hardly alone. Generals and members of Congress were making the same arguments..." The fact that all this is getting swept under the rug in this article, as well as in the obviously habiographic bibliographical treatment that it references, doesn't bode well for the future of scholarly historical research in the USA. Past treatments were far more balanced.
- jongorer 4y agosounds like you're mad because you don't understand simple game theory (and perhaps also jealous of his mental capacities and/or status?). Mr Neumann wasn't advocating for "genocide" (which is a completely false characterization, if not a hysterical one) as much as he was trying to save the world from much worse fate.
- hutzlibu 4y ago"as much as he was trying to save the world from much worse fate." So do you think the world would be in a better shape today, if the world would have listened to Neumann and had its nuclear war? The characterization of starting a nuclear war as genocidal is very acurate, no matter how "noble" the intentions were. Stalinist sowjet russia was hell on earth, but nuking them would have just made hell spread and worse. Von Neumann was clearly a genious. But when you leave out the dark details, it just becomes cult like glorification. I don't think, there should be a disclaimer everywhere he is mentioned, but in a praise article like this, it should be included.
- adastra22 4y ago
- eimrine 4y agoDoes Bitcoin algo has any connections to the game theory? I mean not blockchain only but all the thing with prices, set of constants, mining, rewarding etc.
- senda 4y agoeverything has a connection to game theory
- xbar 4y agoIf this article was even a little interesting to you, I strongly recommend The Man from the Future.
- epolanski 4y agoWhat I find interesting of Von Neumann is that he's basically history's last true polymath, an individual able to excel at multiple loosely related fields. After Von Neumann the world might have produced several other geniuses, but the amount of knowledge required to excel in so many different fields is now so huge that I can't think of any such person.