You and Your Research (1986) | Hacker News Reader

You and Your Research (1986)

cs.utexas.edu

132 pointsBrajeshwar33 comments

Richard Hamming would often give this lecture when he traveled to another university. I was in grad school in 1979 when he showed up to give this talk at the University of Illinois (his alma mater). He chose to give his talk in a lounge rather than a formal lecture room, and it was remarkable because he not only spoke but he listened intently to the students. I remember asking him whey he had settled in Monterey California, and he said he had traveled all over the earth but decided there was no place prettier than Monterey. He was wonderfully warm and insightful.

do you remember what lounge he used for the lecture?

I highly recommend "the art of doing science and engineering" by Richard hamming. It's basically a book collection of his essays and history in the field. In fact I think this might be a chapter of it.

I second this -- fantastic book, and cheap too

I received it as a gift from a coworker and it was a delightful read. It was so inspiring that it made me consider seeking more meaningful work though

Did you find, or create, more meaningful work for yourself?

In the book Hamming tasks the reader with creating in some detail a vision of your future. I did do this, but lacked the foresight to effectively find a way that I could shape my work time to be "Hamming" meaningful.

As an example from the book, Hamming tells a story in which he sits with a group of chemists at lunch. He then challenges them to come up with the most important problems in the field. They did so, and none of them were working on those problems. He ended up challenging them to explain why they had not sought to work on those problems, and was no longer invited to the lunch table.

In this case, Hamming's idea of a meaningful career is a really singular one: contribute to solving the most important questions in your given field. This worked out very well for him, and he claims to be an ordinary man who worked hard over a long period of time. For myself, I'm not sure what the important questions are quite yet :)

Agreed, an interesting book. And also, the youtube video of course: https://youtube.com/watch?v=a1zDuOPkMSw

Thanks a lot! I will check it out.

Just listened to episode 484 of the Changelog podcast[0] this week, where Brian Kernighan was being interviewed. I recommend giving it a listen.

When asked for books or material he would recommend, he only mentioned two things; The Mythical Man Month and the transcript of the talk Hamming gave at Bell Labs in 1986 [1].

Having just read it just yesterday, I highly recommend it. What he says applies to any professional career, even outside research.

[0] https://changelog.com/podcast/484

[1] https://www.cs.virginia.edu/~robins/YouAndYourResearch.html

I'm currently an undergraduate, and I can't shake the feeling that I desperately want to go into academia. I think I truly want to, but just can't convince myself that the relative financial situation and difficulty is worth it. Perhaps it's this doubt that's going to ensure that I don't do well, even if I make the jump into a graduate school.

I was in your shoes (undergrad class of 2015). My PhD journey was way more isolating, cynicism-inducing, political, yet also more successful than I could have imagined. I now have a nice researcher job in industry that has lots of agency, is interesting, and I have no complaints on pay. Yet I also have come to understand that there's nothing special in the intellectual achievements one can have in science. The thought of challenging myself intellectually and striving to reach heights was appealing to me in undergrad. While it might be easier to think critically/scientifically in science vs other areas due to job descriptions, many people think scientifically in other areas with just as much intellectual rigor and insight and without any advanced degree.

Can I ask what company (especially if non-obvious, i.e. MSR, Google)? Currently debating leaving a faculty position for industry research.

I work on ML+bio/chem at Genentech. I interviewed someone who deferred their asst. professorship and I think we're likely to hire them.

You can always start, and if upi don't like it after a year, leave. Most people I know, who did PhDs, knew at that point. Some chose to stay anyway for the sunk cost fallacy, or because they had a visa, but they probably shouldn't have.

Personally I loved every day of doing a PhD (admittedly I did it in Europe, not the US) and I feel like it's allowed me to work on much more interesting stuff afterwards as well.

How good or bad a PhD is depends on the field. I think CS right now is probably a good field to get a PhD in, at least if in a subfield that will get you a job after graduation. I think that's probably the only place getting a PhD would be worthwhile (from a financial perspective), and it may not be that way for long. I've read that economics used to be a field that had good job prospects outside of academia after graduation, but more recently I've read that isn't the case any longer.

I think the incentives in academia tend to lead to overproduction of PhDs in the long-term, as professors are incentivized to produce PhDs but there is no comparable incentive to help PhDs get jobs. Don't be fooled by "I'll write a recommendation letter for you" as that's probably not going to help you get a job. And I do know that NSF requires grant proposals to have a professional development plan, but the plans I've seen tend to be vague and frequently describe things that won't happen.

Keep in mind that you don't need to be in academia to do research. Instead of getting a PhD, I think it would have been better for me to get a real job with real pay for a while, doing research on the side. At some point I'd feel comfortable to switch to a part-time job or stop working for a while so that I could dedicate more time to research. When I was a PhD student, only rarely was I able to give my full attention to my research. I've had a variety of funding: fellowship, grant, teaching assistant, and even self-paid. The only time I was able to focus on my research was when I paid my tuition myself. (When I was on a fellowship, I had classes to distract me.) Grant projects can be hit or miss: If you like the project, it can be good, but if you don't like the project, then it's just a poorly paid job. Teaching assistant positions can require a lot more hours than are implied, unfortunately, and some professors are jerks. If my attention is going to be divided, at least I could get paid decently for it. I do think that by doing research on the side, I would have spent less time on my research, but it still would have been a lot of time over the years.

Can I ask what your subfield is? My impression, outside of CS, is that if you are in academia, you get 1/2 to 3/4 of your time to devote to research (if you get a non-adjunct job), but if you are in industry, you have to do a full-time job before you get to spend time on your research.

My research is in fluid dynamics, though my day job does not involve research.

How much time someone in academia can devote to research depends a lot on their position. Your average professor seems to only dedicate around one day per week to research in my experience (certainly far less than 1/2 to 3/4 of the time), and mostly acts as a manager and a grant writer. (Some people seem to consider grant writing to be research time, but that doesn't make sense to me.) That's on top of the fact that professors tend to work a lot more than 40 hours per week. Around 60 hours per week seems about right in my experience.

There are more research-oriented roles (like postdocs) after getting a PhD, but they tend to be rarer. And I think this is true in academia broadly: I recall when speaking to Scott Aaronson before that he said since finishing his last postdoc, he's never done anywhere near as much research as a fraction of his time. (He did also say that being a parent takes more time than the non-research duties of a professor.)

In my experience, industry (and government) tends to have less hours of work per week. There certainly are research-focused jobs in industry and government (for example, scientists at NASA). But I've heard that in many industry positions, what you said is accurate: Any research one does is after the full-time work is done. To be honest, that might leave you more research time than a professor gets.

If it's something you really want to do, there's no harm in trying it.

You'll develop all sorts of skills that will serve you well in industry should you decide academia isn't for you.

And the opportunity cost of not working in industry those few years will be a drop in the bucket in the long run.

Remember that a career can span forty years. You have time to do more than one thing. Becoming a domain expert skilled at research and mentoring young people is not a bad way to start a career.

>The first person to produce definitive results generally gets all the credit. Those who come in second are soon forgotten.

'Generally'. In more than a few cases, the first person did not receive any credit at all, for years, decades or even centuries. (E.g. who invented the light bulb? or antisepsis? or the transistor? or LEDs? or the first electronic organ? *) Studying those cases may lead you to tactics to prevent being 'robbed'. Several of those 'first' persons were 'inspired' (ahem) by forgotten/obscured work.

* Ever heard of Winston E. Kock?

Actually in fact it's so common there's a term for it:

https://en.m.wikipedia.org/wiki/Stigler%27s_law_of_eponymy

In my own area, I've run into things being rediscovered multiple times.

Man who discovered gfp (green fluorescent protein) was forgotten. Robert Tsien and Osamu Shimomura, and Chalfie got the nobel prize for it, not Douglas Prasher.

The most recent example would be Virginius Siksnys. He did not get the Nobel prize for crispr but it was Doudna and Charpentier who got it instead.

Oh dear, yes. I read the biography of Jennifer Doudna by Walter Isaacson. It's a great book, but I really felt sad for Siksnys. Just because their paper got rejected by Cell does not mean that the paper came after Doudna and Charpentier's - after all, it was submitted earlier than Doudna and Charpentier's. The book overall was not very inspiring with the amount of persnickety politics - but then, perhaps, that's how science is, in the day-to-day grind.

Also there is precedent for Nobel-level papers coming in within months of each other and still be considered for Nobel prize.

See circadian rhythms with Michael Young and Michael Rosbash. They both got Nobel prizes, but one of the papers are several months after the other.

In the gfp case the three winners also dis their work independently if I am not mistaken.

The Siksnys case was discrimination against unknown institutes from unknown countries plain and simple.

The most recent list of all HN You and Your Research episodes from the thread of last one:

https://news.ycombinator.com/item?id=28323486

That’s a long list!

Hamming gave a this talk again in 1995 and it was recorded!

https://www.youtube.com/watch?v=a1zDuOPkMSw&t=134s

I think this is really good. Somehow after reading this one would have expected hamming to have achieved more. Maybe he was missing something in the essay.

> would have expected hamming to have achieved more.

This must be the understatement of the week, considering he was one of the first recipients of the Turing award :)

> I think this is really good. Somehow after reading this one would have expected hamming to have achieved more. Maybe he was missing something in the essay.

This sounds to me like a good advisor. A good advisor can also be a good and accomplished researcher in their own right, but there is no implication either way. (But I also think we shouldn't be too quick to dismiss the work Hamming did—that he wasn't a high-flying rock star doesn't mean we should discount his work.)

Finding the right problem to work on is probably the most difficult part. Renaming problems in most fields are either too difficult or too boring.

Previously on HN:

https://hn.algolia.com/?dateRange=all&page=0&prefix=false&qu...

I recently came across You and your research from The Changelog podcast - https://changelog.com/podcast/484. That episode was so good. By the way, you and your research is a talk at a university by Richard Hamming. You can find it in YouTube.

Also another similar podcast episode was BSDNOW 455: Ken Thompson Singularity - https://www.bsdnow.tv/455. The interesting part about Ken Thompson is a small portion if I remember it correctly.

Really good episodes.

Reply on news.ycombinator.com