UC Berkeley research fellow Michael Cohen dies at 25
dailycal.orgIf anyone's interested, I found video of Michael Cohen as a guest lecturer: https://www.youtube.com/watch?v=n3umLs_SnhQ
How does one publish 30 papers in 3 years? That is an insane rate!
Just navigating the political ramifications, dealing with revisions, etc, on three or four a year is a serious pain in the ass, let alone ten.
Well co-author. First author is one thing, co-author is another. And I no way mean any disrespect to someone who obviously has contributed a great deal to his field.
That's not always true at all and varies greatly from one sub-field to another. I have read countless top-conference papers where the authorship are obviously attributed in alphabetical order.
I don't think there's the concept of first author in theoretical CS? All the papers I've seen there follow the same practice as in Math, where authors are just alphabetical.
No matter how they are listed, someone is probably doing a disprortionate share of the team wrangling and managing the process to publication.
That's another matter entirely. Yes, authors don't always bring equal contribution and effort into an academic work. But, here it's someone claiming that not being first author is a major deal, or bears any significance at all, when in fact it doesn't.
I read the claim that not being first author (who in may, but not all—IIRC, e.g., biosciences are different—other fields is normally the person who is the administrative point) means less irreducible management overhead limiting total throughput, in response to a comment indication that the shear administrative overhead of the number of publications seemed prohibitive.
My concern was not about the administrative burden of publishing but of the difficulty of actually finding results which are publishable.
I'm not sure which the original comment though.
What is considered a paper in CS is different than in many other fields. That is not to say the content is not worthy of publication, just that they tend to be shorter and often are more focused on a single issue. There are also publications in conference proceedings that are usually an easier review process. Looking at his top 10 by citation count on google scholar shows most in proceedings and under 10 pages. Arguably it is more intellectual work, but the process is less arduous so the rate isn't that insane.
"There are also publications in conference proceedings that are usually an easier review process." this is not really true; in systems conferences, for example, you might get 7 or even 10 long, detailed reviews about your work; most papers are rejected (say 80-90% out of 300); a shepherd ensures that if accepted, the paper is appropriately revised. And, it's different in most every subfield of CS. So why make the overly broad (and wrong) generalization?
> There are also publications in conference proceedings that are usually an easier review process.
The big CS conferences are around 20% acceptance rate, Science and Nature are around 7%. So while easier than Science, that is still quite an accomplishment!
Getting a paper into high-impact conferences are definitely still an incredible accomplishment, but comparing these acceptance rates is a bit misleading. Those CS conferences have a very narrow scope, while Science and Nature literally cover any topic pertaining to science and nature... including CS :).
less arduous is a bit subjective no? different types of intellectual labor different types of standards etc including length.
I'm a friend and colleague of Michael's at MIT. This article is very nice and gives a good summary of what made Michael so special. In a lot of ways, Michael was the animating "spirit" of the MIT theory group. He had an encyclopedic knowledge and incredibly deep understanding of basically every area of computer science, and many areas beyond; in the short year that I knew him, we had conversations about everything from convex optimization to computer architecture to rent control laws to Medieval philosophy.
Michael was a truly remarkable researcher. Ludwig's comments about him being the type that you "only see a couple of times in a generation" are accurate. I also recommend watching the start of Yin-Tat Lee's recent talk [1] at the Simons Institute. Yin-Tat is a prolific researcher himself, so his comments carry a lot of weight.
For those wondering about Michael's publication count: computer science (and especially theoretical computer science) is a "high publication" field, in part because of the nature of publishing in conferences and in part because the field is young and there are many good open problems. Still, Michael's publication record is abnormally strong and reflects his collaborative nature. Regarding the comments about co-authorship, Michael could easily have been a co-author on a dozen more papers if he had cared, since he often contributed the main ideas to projects that he never formally joined. This was definitely my experience collaborating with him. I expect that Michael will be more prolific in the next year than many living researchers, from the point of view of publishing.
His papers (incomplete list here [2]) are very well written, by the way. I recommend checking them out.
The most incredible thing about Michael was the way he learned. If you talked about something that he didn't understand, he'd quiz you about it until he did. And he did this with everyone, from brand new grad students like me to famous professors.
At the same time, Michael was incredibly generous. He liked to talk, and you could interrupt him at any time and he'd explain everything to you with astounding patience. Michael wasn't in science for glory; he just really loved learning and teaching. He's already profoundly missed and our entire community is shocked by his untimely passing. My deepest condolences go to Michael's family.
We hope to have a memorial website up soon, especially since Michael was too humble to have much of an online presence.
[1] https://www.youtube.com/watch?v=6pIheZseT1U
[2] https://scholar.google.com/citations?hl=en&user=t3kDJHQAAAAJ...
Life is so precious. To think any one of us reading this could be gone tomorrow.
I'm morbidly curious what the odds are that someone reading this will be gone tomorrow. I'll bet it's non-trivial.
I think you are right: the chances aren't extreme, but are non-trivial. This looks to be a chart that has the mortality data necessary to answer your question (although presumably it also includes "non-natural" causes like suicide and homicide): https://www.ssa.gov/oact/STATS/table4c6.html. The first two columns show age, and the probability that a holder of an American Social Security card who is of that age will die within the next year. For all but the oldest ages, I think it should be accurate to simply divide this by 365 to calculate the chance that an individual of that age will die within 24 hours.
From the chart --- and oversimplifying HN demographics by assuming they match those of chart and consist only of 30-year-old American males --- the probability of death within the year is 0.0015; making each individual's daily chance of death 0.000004, which is about 1 out 250,000. Unless I'm misremembering, to calculate the chance that at least one person out of a group of size N dies, it's easiest to exponentiate the probability of survival (1 - .000004 == 0.999996) ^ N, and then subtract this from 1 to find the chance of at least one death.
If we guess that we readers are one-thousand 30-year-old males, I think that means there is about a half a percent chance that one of us will die before tomorrow ((1 - (0.999996 ^ 1000)) == 0.004). If we instead assume ten-thousand 30-year-old males, then we get about a 4% chance that someone won't be around after tomorrow. If we generously assume a hundred-thousand such readers, then there is about a 30% chance that one of us won't make it another full day. I don't know what the actual readership numbers are for this post (or maybe the grandparent was self-referencing their own comment rather than the main post?) but it seems likely that it's somewhere within that range.
If we use a more realistic age distribution for HN, the probability would go up (older readers increase the probability more than younger readers decrease it). On the other hand, if we assume that HN readers on average have better health care and less risk of violence than randomly chosen Social Security card holders, then the probability would go down. But suicide risk would probably go the other way, so I don't know what the total correction factor would be. Still, I'd guess this estimate would remain in the ballpark. Corrections to my methodology or calculations appreciated.
Nice analysis. One shortcoming is you group all deaths together; expected and unexpected deaths. A considerable chunk of 30 year old's may be expected to die, mostly from cancer.
Given that HN's demographic skews heavily towards young males with high-income jobs, the chances of death tomorrow are probably more slight than for most general audiences.
The odds that a Hacker News user will die tomorrow approaches certainty. The probability that that person has read this particular comment are considerably lower, as every user doesn't read every comment.
I've been playing with the "actuary.py" program and some generated data using awk to come up with estimates based on some very crude assumptions: that there are 100k users, that the ages are uniformly distributed from 20 to 50, and that they are 80% male.
That last number tells you that there's a 95% chance of someone dying within 0.01 years, which is to say, 3.65 days.time ~/bin/actuary.py $( gawk 'BEGIN { srand(); for( i=1;i<100000;i++ ) { age=20 + int(rand() * 30); sex=(rand()>0.8); if( sex=="0") sex="m"; else sex="f"; printf( "%s%s ", age, sex) } }') There is a 5% chance of someone dying within 0.0 years (by 2017). There is a 50% chance of someone dying within 0.0 years (by 2017). There is a 95% chance of someone dying within 0.01 years (by 2017).The script fills out the likelihood that we will all die (we will) ... within a given time:
The actual daily uniques were 300k, and about 3m monthly, as of about 3 years ago. My 100k population may be a good estimate of the actual number of humans in the daily read. The age distribution almost certainly skews generally younger, but extends older, so don't take the values as gospel, only a general indication.There is a 5% chance of everyone dying within 85.66 years (by 2103). There is a 50% chance of everyone dying within 87.7 years (by 2105). There is a 95% chance of everyone dying within 90.71 years (by 2108). Probability of all dying in 1.0 year: <0.001% 100.0 Probability of a death within 1.0 year: >99.99%I also believe the mortality tables Randall uses have been updated since.
The script takes over 6 minutes to run on a rather modest system at 100k individuals.
HN user data (from ~3 years ago):
https://news.ycombinator.com/item?id=9219581
xkcd "actuary.py"
https://blog.xkcd.com/2012/07/12/a-morbid-python-script/comm...
How does one die of natural causes at 25?
Natural causes just means ruling out of external causes and accidental death.
I'd only made the acquaintance of Michael through a mutual friend. In but a few hours, I became convinced this person is one in a billion. What a pity.
Live as if you were going to die tomorrow, learn as if you were going to live forever – Gandhi
How does someone dies of "natural causes" at 25? This sounds more like an unknown health condition.
Over at Scott Aaronson's blog his father has mentioned that, according to the coroner, he may have had undiagnosed type-I diabetes [1]. His father also mentions that Michael supported the charity givedirectly.org if anyone prefers to offer condolences that way.
It's a sad event for the theoretical CS community. I didn't know Michael, but looking at some of the other memorials some prominent researchers have written ([2], [3]), he appears to have been an extremely promising researcher and collaborator.
[1] https://www.scottaaronson.com/blog/?p=3468#comment-1746556
[2] https://blogs.princeton.edu/imabandit/2017/09/28/michael-b-c...
Michael's father posted a comment to Scott Aaronson's blog where he confirmed it was not suicide or homicide, and speculated it might have been undiagnosed type 1 diabetes (among other contributing factors), but said they won't know for sure until the autopsy report is completed.
Preexisting congenital conditions. There is always one or two young people in any given year that decide to do some endurance event and simply drop dead from what looks like a heart attack.
I had two friends die of heart failure, one at age 15 and one at 25. Both healthy otherwise. One died after a soccer match, the other in his sleep one night.
Very sad news about Michael. My heart is with his family and friends.
dying from a disease is still a natural cause. Unnatural would be getting killed, suicide, accidents etc.
The term "natural causes" is a misnomer for "non-human-agent causes".
Not quite, 'Eaten by shark' for example is not usually lumped under natural causes. Even if that's a natural event.
It fits the "natural versus artificial" dichotomy. i.e. anything is natural if it comes about without human agency.
If I die from a lightning strike while hiking, I'm pretty sure that's not counting as "natural causes" either.
I don't mean to be coy (just trying to be respectful), but you should ask your doctor this. I actually did last year, after wondering what my risks were.