tag:blogger.com,1999:blog-786333285568106173.post8044918874702680853..comments2024-01-25T01:05:59.968-05:00Comments on WebDiarios de Motocicleta: Follow-upMihaihttp://www.blogger.com/profile/11599372864611039927noreply@blogger.comBlogger34125tag:blogger.com,1999:blog-786333285568106173.post-54079103636669512972009-11-16T12:41:31.480-05:002009-11-16T12:41:31.480-05:00A simple Google Scholar search of your name and Co...A simple Google Scholar search of your name and Costis' tells us pretty clearly whose work had more impact so far. Sure, citations are not the only criteria, but are a pretty important one.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-786333285568106173.post-71078276059039496482009-11-14T13:46:53.427-05:002009-11-14T13:46:53.427-05:00"But since you bring up combinatorics, I thin..."But since you bring up combinatorics, I think most people would say the TCS surpassed it a while back, both in depth and breadth."<br /><br />I have seen no one else who exhibit their ignorance so blatantly. Kudos, Mihai. Great entertainment.amnoreply@blogger.comtag:blogger.com,1999:blog-786333285568106173.post-81108442911254359572009-11-12T02:52:14.141-05:002009-11-12T02:52:14.141-05:00"Ask your MIT professors who among them was r..."Ask your MIT professors who among them was responsible for RinaP's exit from the PhD program."<br /><br />Let's see ... Judging by RinaP's coauthors, was it David Karger?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-786333285568106173.post-23087750310483915752009-10-18T09:21:23.790-04:002009-10-18T09:21:23.790-04:00Mihai,your posting was very candid. Unlike many ot...Mihai,your posting was very candid. Unlike many others who bitch in private and put on a civilized face in public, you are not a hypocrite. As far as arrogance is concerned, 95% of TCS folks are **arrogant**. May be 94% of them temper their arrogance in self interest. But in the core they are and will remain arrogant. There are 100's of stories of grad students discontinuing PhD due to the arrogance and inhuman behavior of their advisers! Ask your MIT professors who among them was responsible for RinaP's exit from the PhD program. It is a shame that the same guys consider not suitable for a faculty position.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-786333285568106173.post-61832559447211018262009-10-13T01:05:37.612-04:002009-10-13T01:05:37.612-04:00a quote by Muhammad Ali: "It's just a job...<i>a quote by Muhammad Ali: "It's just a job. Grass grows, birds fly, waves pound the sand. I beat people up."</i><br /><br />Is that Costis or Vinod? ;)BCnoreply@blogger.comtag:blogger.com,1999:blog-786333285568106173.post-1351545362031954382009-10-10T21:01:34.846-04:002009-10-10T21:01:34.846-04:00a quote by Muhammad Ali: "It's just a job...a quote by Muhammad Ali: "It's just a job. Grass grows, birds fly, waves pound the sand. I beat people up."Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-786333285568106173.post-69727579297001210392009-10-07T15:33:31.960-04:002009-10-07T15:33:31.960-04:00Your Machiavellian scheme stands exposed:
you wan...Your Machiavellian scheme stands exposed:<br /><br />you wanted to make UCSD theory look so bad that no one would accept their offers for another 10 years and they would be forced to make you an offer when you are ready to move to a university.<br /><br />Elementary my dear Watson!!Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-786333285568106173.post-52657922322977650312009-10-07T15:25:26.798-04:002009-10-07T15:25:26.798-04:00In fact, I would think 90% of TCS people shudder a...<i>In fact, I would think 90% of TCS people shudder at the phrase "data structure". </i><br /><br />You've been hanging around the wrong crowd. Go through the list of theoreticians at, say, MIT and many of them have published plenty and even written books about data structures.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-786333285568106173.post-58453293775897176362009-10-07T15:12:22.499-04:002009-10-07T15:12:22.499-04:00But since you bring up combinatorics, I think most...<i>But since you bring up combinatorics, I think most people would say the TCS surpassed it a while back, both in depth and breadth.<br /></i><br /><br />This is simply not true. There is far more people working in combinatorics than the whole of TCS (I would venture to guess more that 10 times larger). Moreover, combinatorics is at the heart of many other areas such as algebraic geometry (Schubert calculus), representation theory (Schur functions, theory of symmetric functions),<br />statistical physics (dimers, tilings of Eulidean spaces), mathematical physics (Konsevitch's combinatorial formula for counting rational curves), discrete geometry (arrangements, Coxeter groups),<br />Lie groups (theory of buildings) and not to exclude the little part of combinatorics that TCS people are fond of, namely Erdos style extremal combinatorics. I would guess that even this last part is probably bigger and deeper than most of TCS. <br /><br />Very few people outside a tiny community inside TCS are interested in upper and lower bounds for data structure -- however arcane and difficult it might be. In fact, I would think 90% of TCS people shudder at the phrase "data structure". However, somewhat deep knowledge of combinatorics is essential for a large group of researchers in math (including TCS), physics, statistics, and even biology (phylogenetic trees).Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-786333285568106173.post-63216416493656590382009-10-07T13:35:12.957-04:002009-10-07T13:35:12.957-04:00but rather that the field has not attained suffici...<i>but rather that the field has not attained sufficient depth to have serious grad courses in.</i><br /><br />I disagree. If I want to give a good overview of the main upper and lower bounds in data structures (which is probably less than 1/4 of the whole TCS), it will easily take me 2 semester courses. This is just about giving a flavor of the main topics (but probably not getting to the state of the art in many, but giving pointers to current stuff).<br /><br />I won't comment on the attempts to prove Riemann, since I'm not too familiar with the area. But since you bring up combinatorics, I think most people would say the TCS surpassed it a while back, both in depth and breadth.Mihaihttps://www.blogger.com/profile/11599372864611039927noreply@blogger.comtag:blogger.com,1999:blog-786333285568106173.post-50149703543209739412009-10-07T13:18:23.342-04:002009-10-07T13:18:23.342-04:00rgrig: Telling others that their field is "tr...rgrig: <i>Telling others that their field is "trivial" is not a particularly constructive observation.</i><br /><br />AC: <i>I did not say that TCS is "trivial" -- but rather that the field has not attained sufficient depth to have serious grad courses in.</i><br /><br />Right. That sounds a lot more constructive. Thanks for clarifying.rgrighttps://www.blogger.com/profile/02991214367108471744noreply@blogger.comtag:blogger.com,1999:blog-786333285568106173.post-78952841022262625792009-10-07T13:10:36.558-04:002009-10-07T13:10:36.558-04:00Telling others that their field is "trivial&q...<i>Telling others that their field is "trivial" is not a particularly constructive observation.<br /></i><br /><br />I did not say that TCS is "trivial" -- but rather that the field has not attained sufficient depth to have serious grad courses in. This does not mean that there are no interesting or very difficult problems originating in TCS. <br /><br />The amount of literature and research directions developed towards solving say the Riemann hypothesis is stupendous. Just the books written on the subject will probbaly fill several shelves -- not to mention the time to digest even a tiny fraction of it. In contrast, the amount of literarure in TCS is rather small. This is not unusual for a young discipline. For instance, combinatorics was in such as state a few decades ago before people such as Lovasz, Stanley etc. helped to provide structure to the subject by writing influential books. These books now serve as textbooks for grad courses in the subject. Might be in a few decades from TCS will also evolve into such a mature discipline. But it is not there yet for sure.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-786333285568106173.post-44134508193989058522009-10-07T10:05:55.014-04:002009-10-07T10:05:55.014-04:00As a simple side note:
Of course, our community w...As a simple side note:<br /><br /><i>Of course, our community wastes a lot of time, which something I've always said.</i><br /><br />Most communities waste a ton of time [and money], but their participation rate is high. So, no one person can be blamed for it.<br /><br />The TCS community [to my outside eyes] is pretty small, so you can call people [out|frauds] by name.David Coughlinhttps://www.blogger.com/profile/15859839245975431814noreply@blogger.comtag:blogger.com,1999:blog-786333285568106173.post-12500599989804153682009-10-07T10:00:08.774-04:002009-10-07T10:00:08.774-04:00The ramifications as well as the difficulties of t...<i>The ramifications as well as the difficulties of the Riemann hypothesis is too varied to merit a comparison.</i><br /><br />That's a difference in taste in my opinion. To see how sharply others disagree with your position you might want to take a look at <a href="http://www.math.rutgers.edu/~zeilberg/Opinion57.html" rel="nofollow">opinion 57</a>. Telling others that their field is "trivial" is not a particularly constructive observation. It's perhaps better <a href="http://www.dpmms.cam.ac.uk/~wtg10/zeilberger.html" rel="nofollow"> to try to understand each other</a>.<br /><br />But then again, that would make this blog less fun, for some wicked definition of fun. :)rgrighttps://www.blogger.com/profile/02991214367108471744noreply@blogger.comtag:blogger.com,1999:blog-786333285568106173.post-54522103724593261902009-10-07T09:34:47.035-04:002009-10-07T09:34:47.035-04:00You gave the best job talk I've seen in my tim...<i>You gave the best job talk I've seen in my time at the University of Texas at Austin.</i><br /><br />I saw a (non-public) recording of such a talk and I enjoyed it. So, please, try to record and put online at least some of the talks you will give :)rgrighttps://www.blogger.com/profile/02991214367108471744noreply@blogger.comtag:blogger.com,1999:blog-786333285568106173.post-50547756800508356212009-10-06T21:40:21.660-04:002009-10-06T21:40:21.660-04:00I would not really want to be in a department with...I would not really want to be in a department without systems people. (I have no opinion about AI, since I know close to zero about the field, but I'm sure they're also useful :)<br /><br />I normally have a great time with the system folks, and they have many problems that I like. It's only due to unfortunate lack of time that I never did any serious systems work.<br /><br />Sure, their work may be less "mathematically deep," but they solve very nice problems. My main goal is to solve problem that seem interesting; it doesn't really matter whether the solution is "mathematically brilliant"or not :)Mihaihttps://www.blogger.com/profile/11599372864611039927noreply@blogger.comtag:blogger.com,1999:blog-786333285568106173.post-88675889702702555182009-10-06T20:44:22.842-04:002009-10-06T20:44:22.842-04:00Sorry for so many mistakes in the previous post. I...Sorry for so many mistakes in the previous post. I typed it in a hurry :)Vnoreply@blogger.comtag:blogger.com,1999:blog-786333285568106173.post-892911017634251482009-10-06T20:40:17.133-04:002009-10-06T20:40:17.133-04:00Mihai, I support you. You are a brave guy.
I just ...Mihai, I support you. You are a brave guy.<br />I just have one question for you, and it is upto you to answer.<br />"Do you think that a department should hire a TCS person over lets say AI person or hardcore systems person"?<br />What is think is that even though TCS guys are >>>>> smarter than systems people, departments hire them because they bring money.<br />So, I think (and I am sure) that even though you were better or as good as Costis, his research area had more funding than that of yours.Vnoreply@blogger.comtag:blogger.com,1999:blog-786333285568106173.post-793590895944015282009-10-06T19:56:00.520-04:002009-10-06T19:56:00.520-04:00Unfortunately, the really important problems that ...<i>Unfortunately, the really important problems that makes TCS important (i.e. around P vs NP) are by and large unsolved</i><br /><br />I will agree to the "mostly unsolved" part, but not all the important problems are around P vs NP. (If you know as little about Computer Science as you profess, perhaps you should leave this to the people in the field.)<br /><br /><i>Might as well dabble with game theory than think about P vs NP. On this last point I think you will probably agree with me.</i><br /><br />Of course, our community wastes a lot of time, which something I've always said.<br /><br />But there is also a significant fraction that is not doing trivialities. Perhaps you went to the wrong talks. <br /><br />And if indeed most of the stuff here were trivialities, I would invite you to do a bit of non-trivial stuff, get tenure in CS, and then pursue any problem you care about.Mihaihttps://www.blogger.com/profile/11599372864611039927noreply@blogger.comtag:blogger.com,1999:blog-786333285568106173.post-6716904993326771522009-10-06T19:36:46.867-04:002009-10-06T19:36:46.867-04:00The definition of the Reimann hypothesis can be gr...<i>The definition of the Reimann hypothesis can be grasped in less than a day. Now go solve it!<br /></i><br /><br />This is a completely misleading analogy. The ramifications as well as the difficulties of the Riemann hypothesis is too varied to merit a comparison. Having listened to innumerable talks in which TCS speakers talk interminably about things like duality between Voronoi diagrams and Delaunay triangulations (or some such fact deserving a mention but not much more), I must say that TCS people are really fond of extolling trivialities at the expense of the real stuff. And there are real difficult problems in that area -- understanding the level sets of line arrangements, for instance, but there are many others.<br /><i><br />n my experience, mathematicians are uniquely ineffective at solving TCS problems, except problems that are created by other math-leaning people and have limited meaning in TCS.<br /></i><br /><br />Of course, you can define your version of TCS and its important problems to anything you want. Unfortunately, the really important problems that makes TCS important (i.e. around P vs NP) are by and large unsolved -- and it seems to me that only people who have the desire to make progress on these problems are mostly in math. Might be this has to do with the funding as well as tenure structure in TCS which discourages taking such risks. Might as well dabble with game theory than think about P vs NP. On this last point I think you will probably agree with me.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-786333285568106173.post-37344436590518688592009-10-06T14:11:08.604-04:002009-10-06T14:11:08.604-04:00The definition of a Voronoi diagram and its basic ...<i>The definition of a Voronoi diagram and its basic properties are all rather trivial. There is no real added advantage of knowing them before-hand</i><br /><br />The definition is simple, but the important results concerning it (such as, e.g., the fact that it has worst-case complexity O(n^ceiling(d/2)) for any Euclidean space of fixed dimension d, or that it can be constructed in time O(n log n), or that it together with point location data structures can be used to provide an optimal solution to the planar post office problem) are not so trivial.<br /><br />More to the point, though, if you don't already know about Voronoi diagrams, you won't know that they're the perfect tool you need to solve your own pet problem, and no amount of looking-up skills will save you.D. Eppsteinhttp://11011110.livejournal.com/noreply@blogger.comtag:blogger.com,1999:blog-786333285568106173.post-59104814213571163642009-10-06T12:41:35.104-04:002009-10-06T12:41:35.104-04:00Please solve this problem http://maven.smith.edu/~...Please solve this problem http://maven.smith.edu/~orourke/TOPP/P63.html#Problem.63 for us--unworthy TCS folks.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-786333285568106173.post-14800696800681741422009-10-06T11:41:20.151-04:002009-10-06T11:41:20.151-04:00The definition of a Voronoi diagram and its basic ...<i>The definition of a Voronoi diagram and its basic properties are all rather trivial.</i><br /><br />The definition of the Reimann hypothesis can be grasped in less than a day. Now go solve it!<br /><br />If you have a free weekend during this process, can use use Saturday to learn about an important problem in TCS and Sunday to solve it?<br /><br />In my experience, mathematicians are uniquely ineffective at solving TCS problems, except problems that are created by other math-leaning people and have limited meaning in TCS.Mihaihttps://www.blogger.com/profile/11599372864611039927noreply@blogger.comtag:blogger.com,1999:blog-786333285568106173.post-68332129074318039832009-10-06T11:20:38.612-04:002009-10-06T11:20:38.612-04:00when I asked who knew what a "Voronoi diagram...<i>when I asked who knew what a "Voronoi diagram" was, a few sporadic hands came up. (Go try this at home!) The problem: Weizmann is an excellent school, educating many of our future colleagues.<br /></i><br /><br />The definition of a Voronoi diagram and its basic properties are all rather trivial. There is no real added advantage of knowing them before-hand since any grad student or researcher can understand its definition in less than two minutes. Details about complexity and algorithms for computing it might take a little longer, but a mathematically sophisticated person can more-or-less understand everything there is about them in CS literature in a day perhaps.<br /><br />Compare that to say notions from real analysis or probability theory or even combinatorial theory -- all of which would take more than a year to understand at the level needed for them to be useful in research. eg. the notion of Hardy-Littlewood maximal functions (found in the beginning of grad level real analysis textbooks). <br /><br />There aren't really anything in "core TCS" that needs to be taught at the graduate level. If one has sufficient<br />mathematical maturity these stuff are all rather trivial. If one does not have the maturity, then they look ad hoc and impossible to master. Thus, it is much better if one opts towards attaining mathematical maturity first before one starts working in TCS.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-786333285568106173.post-85957358188697071842009-10-05T23:10:58.939-04:002009-10-05T23:10:58.939-04:00With regards to the "anonymous" poster w...<i>With regards to the "anonymous" poster who says he was a board which decided not to interview you, I have mixed feelings about his authenticity.</i><br /><br />I think the comment is genuine. I have friends on these committees, and many of them told me about opinions of this sort. So yes, these committees do read blogs :)<br /><br /><i>Ok so here is what I haven't yet understood, and it is your response to UCSD. It came across as "you did not like being judged second by UCSD and hence you sort of got back at them".</i><br /><br />As I said, there are details that I cannot describe on a blog. Ask me over a beer some time.<br /><br /><i> Now you "praise" (?) Costis but you also write [...]</i><br /><br />Are you telling a scientist that relating some empirical observation is not polite? :) <br /><br /><i>Only small people attribute the other people's successes to luck/politics and their own to their abilities. Do not be one of them.</i><br /><br />While you may be happy to believe that, any cursory reading of a book on the history of science may paint a more nuanced picture.<br /><br /><i>Can you offer any advice on how to save UT?</i><br /><br />I am flattered that you may ask me... My experience at steering departments is quite limited :)<br /><br />Seriously, the hard part is often above the TCS level. If you have the ability to hire 3 top theorists in 3 consecutive years, you will certainly end up with one of the strongest departments around (especially since you build a lot of excitement over your department). But university politics makes this phenomenally hard.<br /><br />With more limited resources, I would guess the best strategy is aggressive student recruiting. There are many students that do not get into MIT or Berkeley for the wrong reasons. If UT applies some non-standard filter and gets a few brilliant students that are falling through the cracks, it will also be wildly successful.Mihaihttps://www.blogger.com/profile/11599372864611039927noreply@blogger.com