5th HLF – Interviews with mathematics and computer science laureates: Madhu Sudan
This is a modal window.
Das Video konnte nicht geladen werden, da entweder ein Server- oder Netzwerkfehler auftrat oder das Format nicht unterstützt wird.
Formale Metadaten
| Titel |
| |
| Serientitel | ||
| Anzahl der Teile | 49 | |
| Autor | ||
| Lizenz | Keine Open-Access-Lizenz: Es gilt deutsches Urheberrecht. Der Film darf zum eigenen Gebrauch kostenfrei genutzt, aber nicht im Internet bereitgestellt oder an Außenstehende weitergegeben werden. | |
| Identifikatoren | 10.5446/40118 (DOI) | |
| Herausgeber | ||
| Erscheinungsjahr | ||
| Sprache |
Inhaltliche Metadaten
| Fachgebiet | ||
| Genre | ||
| Abstract |
|
00:00
Elektronisches ForumComputerMathematikFormation <Mathematik>Computeranimation
00:16
Elektronisches ForumNevanlinna-TheorieDifferenteProzess <Informatik>OptimierungZweiInteraktives FernsehenKontrollstrukturMusterspracheQuick-SortJSONXMLUMLBesprechung/Interview
01:11
Nevanlinna-Theoriet-TestDifferenteUmsetzung <Informatik>BitMengenlehreDickeMultiplikationsoperatorProzess <Informatik>Quick-SortFakultät <Mathematik>Interaktives FernsehenDivergente ReiheFlächeninhaltInformatikMathematikRichtungOrdnung <Mathematik>GradientMultiplikationFormale GrammatikDatenfeldRückkopplungPhysikalismusKollaboration <Informatik>Rechter WinkelBesprechung/Interview
05:48
Nevanlinna-TheorieRückkopplungPunktMAPt-TestPerspektiveStrategisches SpielMengenlehreCASE <Informatik>FlächeninhaltMultiplikationsoperatorFrequenzResultanteUmsetzung <Informatik>MereologieAbstandStatistische HypotheseVerschlingungBitSoftwareschwachstelleBesprechung/Interview
09:01
Nevanlinna-TheorieLokales MinimumAuflösung <Mathematik>MAPQuick-SortQuaderTermKanalkapazitätAusdruck <Logik>Multiplikationsoperatort-TestProzess <Informatik>MathematikerinFrequenzBesprechung/Interview
11:39
Nevanlinna-Theoriet-TestNeuroinformatikProzess <Informatik>EinsNachbarschaft <Mathematik>MultiplikationsoperatorFormale SpracheComputervisualistikMathematikDatenfeldMengenlehreTermPerspektivePhysikalischer EffektSchaltnetzSoundverarbeitungTelekommunikationPhysikalische TheorieRechnernetzÄhnlichkeitsgeometrieGruppenoperationPunktBesprechung/Interview
15:11
Nevanlinna-TheorieComputerspielTeilmenget-TestFlächeninhaltDatenfeldMereologieFeuchteleitungArithmetische FolgeMultiplikationsoperatorWellenpaketBeweistheorieFormation <Mathematik>InformatikMathematikTheoretische InformatikShape <Informatik>OrtsoperatorOrdnung <Mathematik>EvoluteInstantiierungGüte der AnpassungEinfach zusammenhängender RaumBesprechung/Interview
Transkript: Englisch(automatisch erzeugt)
00:22
Is this your first HLF? No, this is my second HLF, though I came to the first HLF and then there was a break of three years and now I'm back here. You were at the very first one? At the very first one, yes. It was very impressive. Have you noticed any differences between then and now? Not yet. I am told that there will be some differences in the programming. However, I think so far it's off to a similar start.
00:47
The first one was very well organized and so is this. So I don't see that there's a major difference yet. I think some of the technical ingredients are changing and the exact interaction patterns are being sort of honed.
01:03
So I'll probably notice some differences as I go further up. But you came back, so apparently you like it. Yes, I definitely liked it and the reason I did not come back is not because I didn't like it. It's just always hard to get away and somehow this time I just had to make an effort, commit to it.
01:22
I had to teach during this week, but I managed to find some other people to substitute for me, so this is working out okay. So what do you get from this? Two kinds of things. On the one hand, I do look forward to the interaction with young researchers. I do want to encourage them, give them tips, give them advice.
01:42
Quite often they ask you very interesting questions and it's always very interesting to hear this. A second thing that I do look forward to is a general outreach. I do think some of the research that we do is often conducted behind the walls of scientific journals and so on.
02:01
We publish there, but that's not reaching out to the broader world and I think one of the missions of the HLF is outreach. I appreciate it very much. In fact, one of the best things I liked about the first HLF was a visit to a grammar school over here. So I think I was seeing, I've forgotten, but I think I saw children who were in maybe ninth grade or something and I talked to them about mathematics and I got some feedback saying,
02:25
oh, the teacher whose daughter was herself in the school said my daughter has decided that she's not going to go into theoretical physics but instead going to go into computer science from now on. So that was a win. I did enjoy my interactions with the students then. I think some of this is coming back for that.
02:47
Now, since you teach, you also, I assume, have some students who end up, you become a mentor to them. Right. Can you talk a little bit about what that process is like and how you think it works? It's a, you know, so I've supervised about 20 PhD students and so they are themselves faculty members, many of them now.
03:07
So, yeah, it's a slow process. You are very initially, you spend a lot of time learning about what the students want to do, what they are capable of, what their passions are and what they think their passions are. You try to separate the two.
03:26
You try to separate, you know, people sometimes over ambitious in some directions and under ambitious in others. And you try to sort of try to train them into saying, well, you know, these are kinds of things that you are actually very good at. Why don't you do more of this? But the process is definitely, I mean, it's almost like raising a child.
03:46
It's a typical PhD student spends about five years under your supervision. They're not working exclusively with one faculty member, not in my field. They would often talk to multiple people with different people.
04:01
You have to have different length ropes, different length leashes and say, you know, you should really just wander off on your own and do whatever you want. Some people really want to wander off and you want to say, no, no, I want to hold you on a tighter leash to make sure you do some of the things that you're supposed to do. You spend, I think, a bulk of the time worrying about how they're going to perform.
04:24
And the last year, absolutely thrilled with how they're performing and say, finally, I found a collaborator and this person is going to be great to work with. And then they're ready to graduate and they go away. It's very sad. Well, comparing that to the HLF, you only have a week here with these people.
04:40
That's right. Yeah. So I think even here, I was actually just observing to a colleague over here that, you know, first day, all the interactions are very tentative and we sort of, you know, are feeling each other out to see, you know, is this, especially the young researchers are thinking, is this a person that I want to talk to? Is this in my area? Will they be able to tell me anything interesting? And
05:04
as time passes, the second day, the third day, the boat trip especially helps a lot. You spend a lot of time with them. Over time, they start to realize that, you know, they find the right people to talk to and you often have multiple conversations with the same set of people. I vividly remember last time we went up to the castle and we were walking on the terrace and
05:26
there was just, you know, a series of wonderful interactions, one after the other, rapid, but very, very, very, I think, informative and useful and not just feeling each other, where are you from, what are you doing, but really more like, what should I do if I, you know, if I'm trying to do this and I'm stuck?
05:43
And you say, look, why don't you try the following approach? And, you know, the approach might be talk to person X or go look online, etc., etc., but, you know, for certain things, but these are very, very concrete pieces of advice. And they actually, I mean, they make sense to me. I did not get major feedback afterwards saying, look, I actually did this and this happened.
06:05
There were a few students who went back to their home institutions and wrote to me saying, you know, what about the papers that you spoke about? Where are the links? How can I find out more? Some were even interested going a few steps beyond, but, you know, as with all the other things, you know, after a while, there's the usual attention level decay on my part.
06:25
You know, two weeks into the thing, I'm able to pay some attention, four weeks later, it's a little bit less. After a while, the conversation, the speed of the conversation decays to the point where it's almost zero. So it does end after a little bit, but I think it does have, you know, it does have influence,
06:44
a lot of influence on a small set of people, some broader influence, I would suspect, on a larger set. So think back to when you were around this age, 22 to 27 or so. What were you doing? What was that five-year period like for you?
07:02
So during my PhD, I went through some fairly vulnerable periods of time. It would be traditional for a person in my area to have a first publication in their first or second year of some significance and a few more in their third year and, you know, pretty much bursting at the end of the fourth year.
07:26
And I didn't quite follow that usual path and so on. I was at least delayed by, I would think, one and a half to two years and I was going through everything much more slowly. And it was definitely a period where I said, you know, am I going to be of any worth and am I going to be able to find this thing?
07:43
But what I found was very useful was that I'm always very interested in finding, you know, simple explanations for the things that I hear about. And this is a very risky strategy. I mean, you are supposed to go find one problem that you want to work on and try to do the best you can at it, not look at 20 different problems and try to find simple explanations in each case.
08:05
Who cares about that? That's something that may be worthy when you're a senior. At this stage, it was not. But at the same time, I realized that actually this worked out. I mean, things clicked and things worked. And I appreciated the value of my mentors, my PhD advisor, who in fact at one point told me, look, in
08:23
your PhD, you should be writing about these things, not those things, because this is where I think your perspective is relevant. And even as I was writing the thesis, things evolved. I got new results and new things happen. And so it was actually very, very valuable. I think mentorship plays a very significant role. This is why we don't do distance education and PhD stage.
08:44
And at all stages, it matters. I think and I appreciate all the love and nourishment of my research that happened through my peers and mentors. And I'd like to give back in the same way.
09:02
It's interesting you talk about this vulnerable period, and I imagine some of these researchers are going through it. What advice would you give to somebody who's in that period? So I think the first thing that we try to tell, we do it sort of in a cautious but gentle way to make sure the students' passions are really aligned well with the world of research.
09:26
If you're passionate enough, I think you should just go ahead and pursue things and you should do things that you're good at. You should not do things that others tell you to do. So if somebody tells you, look, the right way to do it is to go read 20 different papers and try to find simple explanations, that person's probably wrong.
09:43
That person is probably good for themselves, but not for you. On the other hand, if somebody else comes and says, okay, you're looking at these very deep theoretical things, but this is going to be of no practical use whatsoever and this is not going to get you a job. Well, if your interest was in getting a job, then maybe this is not the right path to be seeking.
10:01
If you want to get a good job, there are many, many other paths which are easy to succeed in, but do not have the same level of uncertainty and risk associated with it. Here in the world of sort of, I would roughly characterize myself as a, I'm not quite a pure mathematician, but I align closely with them in terms of my passions and interests.
10:25
I would think we do the kind of work we do because we're very, very passionate about it. Nothing else would be sufficiently satisfactory. When you're at this stage, you really have to go ahead and do what you think is right. People can get you out of local minima. You're stuck in some problem and you're having a hard time struggling out of it.
10:48
Go talk to somebody. That somebody might say, okay, here's something else that you can do for the next few days. But if you really come back to that problem and start thinking about it again in two months, and then again in two months, and again in two months, it probably means something.
11:01
It means you really want to solve that problem. Go ahead, solve it. Don't throw it out. But at the same time, don't box yourselves in. It's always important to make sure you get good stimulation from the outside. You keep your mental capacities well greased and well oiled so that they're constantly thinking, thinking about new things.
11:21
But do not take any formulaic advice from other people either. To me, the most important thing is try to exploit all your natural strengths. Do not try to imitate someone else. It's really good to hear all of these mentoring points, though, because not everybody is as engaged with students as you are.
11:46
It sounds like that's really very important to you. Right, it is definitely. Going back to your own field, what, first of all, have you heard anybody in the two times that you've been here who's caused you to look at it differently?
12:02
And also, what, well, these are two separate questions, I suppose, but what do you hope the researchers here who are in your field will go away and accomplish? OK, the first question in terms of have researchers influenced me to think about things differently? I mean, I definitely got a bunch of questions after my first talk over here, which forced me to think about things slightly differently.
12:28
And it's always a tension between sometimes I think I'm doing the wrong things and sometimes I think I'm explaining it wrong. And finally, there's always an unstated osmosis of ideas. You slowly absorb what the other people are asking and you understand and appreciate it.
12:48
So to me, every exposure does this for me, and to me, HLF is very valuable. This set of people that I meet, the young researchers as well as the senior researchers, they're both unique.
13:02
Each set is unique and definitely the combination makes it even more exceptional. So I've definitely learnt a lot, though I cannot probably pinpoint an idea and say, oh, this is something that I did not do then, that I'm not. Things have been evolving. I'm doing a lot of things now that I did not do then. But I cannot say there's a cause and effect relationship. It probably is.
13:23
Now you've got me curious. What was your first talk and what was the thing that caused you to look at it differently? Both times I've been very interested in, I'm talking about theories of communication. So communication, I think about computer to computer communication, cell phone to computer communication,
13:40
human to computer and human to cell phone communication and human to human communication. I want an overarching theory which can capture everything. And the first time around, I tried to talk about how is it that you would describe the process by which, say, a child is brought up and starts to learn how to communicate.
14:00
And how would you try to build a similar feature into computers? You turn a button on and suddenly starts to acquire language and slowly communicates with its neighborhood. So I tried to talk about it and there were lots of tensions, people saying, why don't you just do X? And why isn't this problem already solved? Or we have computers which do this.
14:22
And so I had to articulate some of the things. It also got me thinking about what is it that I could try to do which doesn't start. Maybe when people think about language acquisition, you can think about what happens with a child growing up. You could also talk about what happens when you move from one country to another.
14:45
Or when you start to talk to a new set of people from a new age group, suddenly there are some linguistic changes, some phenomena. How do you look at the mild changes as opposed to just the entire drastic ones? So all of these different perspectives started emerging around that time.
15:04
I don't know exactly which ones were before and which ones are after and which ones were influenced by the talk here, but some of them were. What years were you an undergraduate? I was an undergraduate between 1983 and 1987 and then did my PhD from 1987 to 1992.
15:24
What challenges did today's researchers face in your field that you think you didn't or that you face that they don't have to? Interesting. On the one hand, I think the fields are getting older. My field is the intersection of computer science and mathematics.
15:43
Computer science was much younger then and it's getting older, so many of the lower hanging fruit are not ripe for picking. The kind of work that I did then used to be considered the deepest mathematical work in my field at the time. Today this thing is probably taught in graduate courses and the kind of work that people are doing on top of it.
16:07
Almost it will require five years of training to understand all the mathematics behind it, not just to explain the single proof. So the field has matured a lot, which makes it harder for people to get into some aspects of the field.
16:24
At the same time, I think my field has become a lot more outgoing. Twenty years ago, the field of theoretical computer science was neither considered a part of mathematics nor computer science. It was really considered an outsider to both fields.
16:42
Today it's almost considered an insider to both fields and part of it is the evolution of our field, part of it is the evolution of the others. A theoretical computer scientist today is considered a lot more valuable resource in a company like Google, Microsoft, Yahoo, in an academic institution.
17:06
Three of my students are now holding joint positions in mathematics and computer science. Most of my predecessors and I have only held positions in computer science departments, for instance. So things have evolved a lot. There's a lot more acceptance of the, a lot more respect for our area.
17:27
So I think the students that are coming in, on the one hand, they're finding the barriers higher. On the other hand, I also feel that they will hopefully find a much better reception to their work, a much broader reception.
17:43
And the problems have changed and that's part of the reason why there's broader reception. People have found the problems which are more relevant, more mathematically challenging, more better connected. And they have been able to make progress on these and this has greatly increased the respect for the area.
18:01
I think the students are working on very interesting and exciting problems. They're coming up with one, they're very talented people in the field. And so I think they, on the whole, I think they're going to be in a lot better shape than we were. It's just that we are occupying many of the positions that they'd like to occupy.
18:23
Well, we know the solution. I don't like that solution that you have in mind. So we are almost out of time. Is there anything else you'd like to say, either about the HLF or the students? I just, yeah, I mean, I'm looking forward to this visit to HLF as well.
18:43
I'm really looking forward to the visit to the school, which is going to happen again on Wednesday. And overall, I'm, you know, I'm really encouraged by the fact that HLF has decided to take this path, you know, to bring the laureates and the young people together. It's good for the fields. I think that's the most important thing.
19:04
The more we manage to put in some, you know, encouraging, more of an encouraging field we can set up for the young people, the more they will thrive and the better it will be for the field, I think. So I'm looking forward to more, five more happy years.