Talk:Main Page

Have a comment? Leave it here (at the top, so that the comments are in reverse chronological order).

A Question

Hi Anton, I think that your situation i.e. going from hardcore pure math to working at google is very interesting and I would be very interested in knowing a little bit more about it. The kind of questions I have in mind are :
Do you feel that your background is helping you at google (if you are allowed to do so it would be very nice if you could explain a little bit what it is you do) and if so in what way ?
Do you miss math and if yes do you try to keep doing some as a hobby (and if so how do you organize your time) ?
Are there many people at google who followed the same path ?
Could you maybe explain why you didn't get a job offer because from where I stand it seems like you have a far better resume than a lot of people that I know and did get a job in academia ?
Basically anything you can say about your experience would be great.

The reason I ask is that I'm considering pursuing a phd in math (arithmetic geometry) myself and I keep hearing that after a phd in pure math you are very attractive to companies such as google but have never heard of anyone having made the switch, so i'm wondering if it is really true. Obviously I'm hoping to pursue a carreer in academia after my phd but I know it is very hard and i'm trying to be realistic.

Anyway I really like all the ressources on you site so thank you.

I'll give some quick answers here. If you want to know more about my situation, or you want to tell me more about yours, feel free to email me.

If your final goal were to get a job at a company like Google, entering a math PhD program is not a very direct way to accomplish that goal. I haven't served on hiring committees or done interviews yet, but I'm pretty sure the reason a math PhD is appealing in an application is that you cannot have gotten one without (1) making the cut with some other selection committee to get into grad school, (2) having a strong capacity for effectively learning and using abstraction, and (3) being persistent enough to finish (this includes willingness to do the less glamorous tasks required). The actual content of your work during your PhD is not likely to be of much interest to a potential employer in industry.

But given that your final goal is academia, the PhD is a must. I don't think it's very hard to maintain an exit strategy in case you can't find academic jobs (or can't find any that solve your two body problem, or you become disillusioned with academia, or whatever). The reasons above for why a PhD is appealing apply for just about any job you are likely to want, so it's certainly not a waste of time to get the degree. Moreover, the things that will make you more appealing to industry employers (e.g. learning to program, or to use a version control system) are also useful if you stay in academia.

To answer your specific questions,

• Do you feel that your background is helping you at google (if you are allowed to do so it would be very nice if you could explain a little bit what it is you do) and if so in what way ?

I'm not doing any algebraic geometry at Google. Maybe I will one day, but it's not likely that the specific expertise I developed in my academic career will be applicable. At first I found this somewhat depressing, but it's not. Even if I'd stayed in academia, working in very different fields would be a sign of success. My move to software engineering is more extreme than that, but not by much.

Right now I'm working on improving the performance of natural language processing at Google.

• Do you miss math and if yes do you try to keep doing some as a hobby (and if so how do you organize your time) ?

Yes, I still like to do some math. I've mostly been learning lots of new stuff, like information theory and machine learning, but I do occasionally chat with my old collaborators about algebraic geometry. I'm really happy that (1) learning new stuff is surprisingly easy, and (2) I don't appear to have gotten much worse at reasoning about algebraic geometry even though I'm not thinking about it regularly.

I'm not very disciplined about organizing my time. I just read, watch lectures, or think about problems whenever I have the time and motivation to do so.

• Are there many people at google who followed the same path ?

I don't think so. Most people I work with have a computer science background, and I think most of the math people come from applied math. One person I've worked with used to be a physicist, but I don't know enough about his story to tell it.

• Could you maybe explain why you didn't get a job offer because from where I stand it seems like you have a far better resume than a lot of people that I know and did get a job in academia ?

I have some speculations, but I don't know for sure. It's hard for me to be dispassionate about it, so I'd prefer not to list all of them. I will mention a couple, with the caveat that I don't know for sure how much of an impact these things made.

For one thing, I was slow to become productive. I spent 7 years in grad school and didn't have any actually published publications at the end of it. My last year in academia was very productive, but none of those papers are actually published yet either (though most of them have been accepted). Other mathematicians write more papers than I did, and they get them through the mathematical publication machine faster than I did.

For another thing, I really like working with other people. I find myself to be much more productive and energetic when I have a partner on a project, so I don't have any single author papers. I think this hurts my application to an academic job, since a hiring committee doesn't have anything that it can be absolutely sure I'm personally responsible for.

Anton 08:51, 14 August 2013 (PDT)

Re: A confession

Thanks! [1]

MATH2SOFTWARE

Hi Anton, I see that you were doing some really interesting stuff in algebraic geoemtry, why the huge shift to software engineering in google?

Hi. I've been meaning to write up the story, and I hope to do so soon. The short answer is that I didn't get any academic job offers, so I applied for non-academic jobs, and Google offered me an excellent one. Anton 20:20, 4 November 2012 (PST)

Comment!

You're using a Wiki, that allows for external contributions, for a personal website. It's slightly logical, but I'm wondering if you actually considered it while building the site.

Your logo is a weasel with expensive taste in wine and ... good bread? Oddly enough, this makes more sense to me then the wiki.

Either way, the site itself is enjoyable, though seldom updated...

Hi! Thanks for leaving a comment. I recently (~6 months ago) converted my personal site to a wiki. The main reason is that I found it was a pain to update my web page, so I would almost never do it. As a wiki, the editing mechanism is sufficiently smooth that I find it easier to update. Though the content doesn't change much, I do find it handy. I've tried it as a sort of quasi-public notebook where I dump ideas I have about some topic (e.g. Properties of morphisms or Artin's criterion for representability), but I don't have a strong argument in favor of this behavior. In part, I just wanted to learn more about wikis and explore the possibilities of the platform.

The fact that it allows for external contributions isn't a problem. Important pages (like the home page, my CV page, or my research page) are protected so that only I can edit them. It did take me several tries to get an effective way to avoid spam (see Working with MediaWiki#Handling Spam).

I can't make much sense of the logo for you. It's a combination of clip art that came with my wife's first computer. For some reason, I started using it as my site logo many years ago, and it stuck.

Anton 16:45, 10 January 2012 (PST)