Follow Slashdot stories on Twitter

 



Forgot your password?
typodupeerror
×
Businesses Space Science

A Growing Number of Astrophysicists Are Leaving Academia To Work For Tech Startups (wired.com) 75

Space scientists are abandoning the heavens to help you decide what to wear and watch and listen to. Whether it's stars or Stitch Fix, it's all about machine learning. Wired: Chris Moody knows a thing or two about the universe. As an astrophysicist, he built galaxy simulations, using supercomputers to model the way the universe expands and how galaxies crash into one another. One night, not long after he'd finished his PhD at UC Santa Cruz, he met up with a few other astrophysicists for beers. But that night, no one was talking about galaxies. Instead, they were talking about fashion. A couple of Moody's astrophysicist pals had recently left academia to work for Stitch Fix, the online personal styling company now valued at $2 billion. Moody gawked at them. "They were like, 'You don't think this is an interesting problem?'" he says. Indeed, he did not. But when his friends described the work they were doing -- sprinkling in phrases like "Bayesian models" and "Poincare space" -- predicting what clothes someone might like started to sound eerily like the work he'd done during his PhD. Quantifying style, he discovered, "turns out to have really close analogues to how general relativity works."

Four years later, Moody works for Stitch Fix too. He belongs to a growing group of astrophysicist deserters, who have stopped researching the cosmos to start building recommendation algorithms and data models for the tech industry. They make up the data science teams at companies like Netflix and Spotify and Google. And even at elite universities, fewer astrophysics PhDs go on to take postdoctoral fellowships or pursue competitive professorships. Now, more of them go straight to work in Silicon Valley. To understand what's driving astrophysicists into consumer product startups, consider the recent explosion of machine learning. Astrophysicists, who wrangle massive amounts of data collected from high-powered telescopes that survey the sky, have long used machine learning models, which "train" computers to perform tasks based on examples. Tell a computer what to recognize in one intergalactic snapshot and it can do the same for 30 million more and start to make predictions. But machine learning can also be used to make predictions about customers, and around 2012, corporations started to staff up with people who knew how to deploy it.

This discussion has been archived. No new comments can be posted.

A Growing Number of Astrophysicists Are Leaving Academia To Work For Tech Startups

Comments Filter:
  • Join an amazing rock band
    • Like Dr. Brian May
    • Astrophysicists can do even better. Half of the Nobel Prize in Physics [nobelprize.org] this year went to two that found an amazing rock: one in orbit around another star. Being in a rock band may get you famous for a few decades or perhaps even as long as a century but the first people who discovered an exoplanet will be known for as long as civilization lasts...although at the rate we are going there is a chance that might be less time!
  • by Retired ICS ( 6159680 ) on Wednesday October 09, 2019 @12:36PM (#59288482)

    Lots of Astrologers are leaving their fortune telling and tarot cards to join startups too.

    • Lots of Astrologers are leaving their fortune telling and tarot cards to join startups too.

      McDonald's is still considered a startup?

    • Lots of Astrologers are leaving their fortune telling and tarot cards to join startups too.

      No, those go get MBAs and make their fortunes the easy way.

    • by Nidi62 ( 1525137 )

      Lots of Astrologers are leaving their fortune telling and tarot cards to join startups too.

      I thought they all went to work for VCs

  • by damn_registrars ( 1103043 ) <damn.registrars@gmail.com> on Wednesday October 09, 2019 @12:37PM (#59288484) Homepage Journal
    This is a very common transition right now. As the abstract eluded to, there are very few academic track positions open for a rather large community of new(er) PhDs. Modern science though handles really large data sets and really complicated models, and those tend to fit with a lot of different industrial goals.

    It doesn't hurt that the money is better, too. A lot better. Postdocs in industry average substantially better pay than mid-level faculty, and they get there without having to work at retail-level salaries as academic postdocs first.
    • Apparently the correct usage of "alluded" eluded you.

    • by Anonymous Coward

      The money is not only better, it's fucking absurd.

      $175k/yr for a typical senior dev? Fuuuck.

      It's a good time to be in code. At least until the next hit and our salaries go back down.

    • It doesn't hurt that the money is better, too. A lot better.

      In astrophysics, yes. In many other STEM fields, not so much.

      Regardless, hopefully these guys are paying into retirement because there’s no tenure in the private sector.

      • by gtall ( 79522 )

        And the probability of landing a tenure track job at a uni is quite small. The states have been cutting higher education for decades, I tend to think of it as the revenge of the MBA and Lawyers. They don't get science or engineering, and everything must turn a profit, hence whack the unis. The federal government is no better. And most of the pols are anti-science, funding it takes away money they could be spending on trinkets to their voters for the next election. And science does bad things for them, it pr

      • It doesn't hurt that the money is better, too. A lot better.

        In astrophysics, yes. In many other STEM fields, not so much.

        I don't know which STEM fields pay well, if any. I'm in biotech and the major research university nearest my home starts faculty in my field at around $75k, with the expectation that they will work around eighty hours per week until they land their first big independent research grant (the very best and the luckiest might land that after about 3 years, others will lose their jobs going for it).

        Regardless, hopefully these guys are paying into retirement because thereâ(TM)s no tenure in the private sector.

        Tenure is quickly becoming a thing of the past. Along with it we see universities that are tying faculty salari

        • Our state employee salaries are all public record so this is easy to look up.

          Full time EE and CS assistant profs are starting above 100K and climb to 150K in 5-6 years. Full professors with research funding are drawing 220K to 350K - those which aren’t funded appear to earn around 150-180K (on a 9 month appointment). Tenure is still a thing here.

          I’m sure astrophysicists are way below that though. And I don’t know how this compares to the private sector jobs these people might have.

          • Our state employee salaries are all public record so this is easy to look up.

            Based on the numbers you post, you likely don't live in the same state I live in. Different universities will have different starting wages for faculty, often based on cost of living.

            Full time EE and CS assistant profs are starting above 100K and climb to 150K in 5-6 years

            I don't immediately know what EE/CS faculty are paid at the major university nearest my house, but I do know what biomedical faculty are paid and they most definitely do not start at $100K here. Assistant faculty here start at $75-80 in anything biomed (including chemistry, biology, biochem, biomedical engineering, medicine

    • Not that this should surprise anyone, but Slashdot clearly still hates unicode. They've been promising to start supporting it for well over a decade now and we see it is still not even good enough to be upgraded to the description of broken.

      What did I run in to now? I had a ">" (greater-than sign) in my subject line originally here. It was dropped. The subject should have been

      Hard Scientists -> Data Scientists

      ... hence indicating a transition of hard scientists into data scientists ... but instead it became

      Hard Scientists - Data Scientists

      Suggesting some sort o

  • Their talent is more useful down here than up there.

    • by gtall ( 79522 )

      Yah, that damn Einstein!! What was he thinking coming up with all that useless theory that allowed us to build the GPS system. If only some bright spark like you had told him to produce the theory so 60 years later the GPS system could be built, he'd have had so much more direction in his life.

  • My oldest daughter began her education in Astrophysics , her boyfriend is still in it.

    She loved physics and space, but despised coding.

    If you want to be an astrophysicists these days you need to code.

    These guys having learned to code, and being intelligent , are a great fit for positions like this.

    • "She loved physics and space, but despised coding."

      This is perfectly normal. Ventricular Fibrillation can be quite painful, and the chest compressions can break ribs. No to mention that defibrillation is also somewhat painful. Most people (who are not masochists) do not like coding either.

      • by kackle ( 910159 )
        Ah! It took me a second, but I got it!
      • "She loved physics and space, but despised coding."

        This is perfectly normal. Ventricular Fibrillation can be quite painful, and the chest compressions can break ribs. No to mention that defibrillation is also somewhat painful. Most people (who are not masochists) do not like coding either.

        Astrophysicists need to know how to write computer code. /wink

    • This reminds me of Walt Whitman's "When I heard the learn'd astronomer" poem. It's interesting how much further astronomy went into the mathematical and coding side of things.
  • by kackle ( 910159 ) on Wednesday October 09, 2019 @01:00PM (#59288638)

    But when his friends described the work they were doing -- sprinkling in phrases like "Bayesian models" and "Poincare space" -- predicting what clothes someone might like started to sound eerily like the work he'd done during his PhD. Quantifying style, he discovered, "turns out to have really close analogues to how general relativity works.

    Yes, but in the end, you are predicting fashion trends. The same trends that are fickle, change all the time, and are possibly mocked by the next generation. Profitable, perhaps, but not very helpful to mankind, I'd argue.

    I've noticed what I would call a big problem lately: The smartest of us can so easily waste our time/lives today. I'm not talking about "having fun" or "zoning out on TV", etc., but rather situations like TFA describes, or the endless electronic device distractions, the constant "news" feeds, the delirious enthusiasm for the latest technology products to buy that don't enhance our lives much at all (or we ignore the downsides of each new thing we pile on) and/or the fact that certain situations have become so complicated, that we spin our wheels frequently dealing with their fallout, when, in reality, the reason for the complication doesn't pan out if anyone bothered to stop and actually think about it. Let me provide a poor example...

    My old car came with a $200 option where the rear-view mirror automatically tints when it thinks it's nighttime and it senses bright lights from behind. All that complication, the electronics, the manufacturing, the pollution and energy behind the electronics' creation and continual use, just to replace the manual flip switch that they've had for several decades. Why? The thing even weighs more, reducing the MPG of thousands of cars. WHY? I also think that someone(s) smart spent much time basically making an expensive mirror.

    There's a large bridge in town at such a position that the mirror (which you can't tint by hand) won't tint when the sunset is blinding from behind because the mirror knows it's daytime! And, after ~ 10 years, the LCD "tint" started to bleed, making the mirror useless. I replaced it with one from the junkyard for $5 which I still have 10 years later.

    I think we could get so much further in a generation or two than fancier websites.

    • This "auto" mirror looks like something that can be done with a couple photo cells and simple logic gates. Very cheap, inexpensive stuff. Of course, the "auto" mirror in reality most likely uses a full ARM based computer system, a camera, and 'AI', and also phones home to upload your personal data to the mothership.
      • by kackle ( 910159 )
        Still, it's more complicated than a manual flip bar, which is exactly my point. (Not to mention, making an LCD from scratch is non-trivial.) Our smarter people should be doing smarter things. We are wasting them, to humanity's detriment.
    • by kackle ( 910159 )
      I meant to say my junkyard mirror was a plain, manually tinting mirror, if I wasn't clear.
    • I've noticed that same problem too. Although... maybe we've just replaced a previous generation's mechanisms for wasting talent (people dying from preventable diseases, spending their lives farming because 90% of the population had to, getting killed in massive bloody wars) with more benign ones (distracted by video games, working hard making complicated things that are profitable but unproductive).

      I'd like to hope that we will mature as a species and learn to actually make good use of our best minds, but t

  • by JaredOfEuropa ( 526365 ) on Wednesday October 09, 2019 @01:00PM (#59288640) Journal
    Remember the narrator talking about scientists focusing on "curing baldness and prolonging erections instead of solving real problems"? This... feels a little bit like that.
  • Scott Manley, noted scottish astrophysicist, and also noted youtube KSP... person? Influencer? Gave up his career in science to go work for a silicon valley FAANG company sometime in the last 2-3 years. Working on.... encoding algorithms? You can make $80-120k as an astro-p at a university, or you can start at $150k at a FAANG, probably closer to $200k with total comp. As a guy with a family it seems like a no-brainer.

  • by littlewink ( 996298 ) on Wednesday October 09, 2019 @01:20PM (#59288756)
    Which shall it be? Shall I keep pursuing my Ph.D (3 more years of slavery studying under my egotistical asshole mentor)? Or should I turn to data science(where linear algebra and a little probability are all the challenges I'll face). I'll make a six-figure salary almost immediately and millions over the next decade. I might even get laid! 8=)) Which is best? It doesn't take an astrophysicist to make that decision!
    • by Shotgun ( 30919 )

      You know what you can buy with a six-figure salary?

      A telescope. And a REALLY good one.

  • Comment removed based on user account deletion
  • The Higher Ed bubble will soon burst. The number of consumers (suckers) for their product has been gradually decreasing. Alternative forms of education that don't leave you permanently indebted are beginning to catch on. The cushy, guaranteed for life jobs will eventually disappear so they are better off doing things that are useful for the market.
    • by habig ( 12787 ) on Wednesday October 09, 2019 @01:54PM (#59288950) Homepage

      The cushy, guaranteed for life jobs will eventually disappear so they are better off doing things that are useful for the market.

      FWIW: I do particle astrophysics, working with a bunch of pure particle physicists. You could do a search and replace: s/Astrophysics/Particle Physics/ in TFA and it would still be true.

      I'll also point out that friends who have made the academia jump from "cushy" jobs to industry jobs making 2x or more the money also report having to work far less hard than they did in academia. They have weekends off! Wow! I'm getting ready to spend my weekend doing research, since teaching fills up the weekdays. Sounds "cushy" to me. Yes, I already have tenure.

      Many people bust their asses on the academia track not for money, not due to indolence, but because they love what they do. Not everyone: but there sure are a lot of us.

      • I'll also point out that friends who have made the academia jump from "cushy" jobs to industry jobs making 2x or more the money also report having to work far less hard than they did in academia.

        Either you or said friends are lying.

    • by gtall ( 79522 )

      Even that Enlightenment was a total wash. Imagine developing a lot of math and theory with no idea they could be working in the village black smithy for much more money. What the hell were they thinking. Too bad you weren't around back them to tell them to stop wasting their time.

    • Anyone paying their own way thru a PhD is a fool and shouldn't be getting one.
    • Alternative forms of education that don't leave you permanently indebted are beginning to catch on.

      This is a US problem that most of other countries does not have, as we have free/accessible education paid by higher tax rates.

    • Yeah, given a choise of having a monster debt in order to become 'enightened' in astronomy or raking in the dough with some vapid tech company, I think I'll be following the money. My idealism died ages ago. It's about the money now. I'm seeing this as the beginning of the end of traditional 'higher knowledge' institutes, both the schools and the workplaces, Not the complete end as we still need professional sismologists, meterologists, and astronomers to keep us safe from killer astroids, but this structu
  • This has been going on for a while now. A friend of mine who was in a physics graduate program over a decade ago talked about how hedge funds raided physics departments to hire the raw mathematical talent as quants. Maybe the only difference is the demand is growing, so the number of people being poached is becoming more and more noticeable.
  • This is an ad for Stitch Fix.

    Nobody cares. The company is not worth $2,000,000,000.00. Idiot investors may have valued it at that. It's probably worth $200,000.00.

  • the college bubble burst is coming but your loan will still be there

  • Pure science has applications. Astrophysics being useful in a fashion app is no weirder than abstract algebra or graph theory becoming incredibly useful once computers came along.

  • ... couple of centuries or so they will look back and call this the Really Colorful Ages.
  • by Tablizer ( 95088 )

    Most scientists go into science for the love of science, not money. If you really dig money, science never was the best bet. But if their science pay grows overly lousy, then tech jobs will beckon them away.

  • Shit, I'd hate to have Sheldon Cooper as a coding partner. [c2.com]

  • by joe_frisch ( 1366229 ) on Wednesday October 09, 2019 @08:04PM (#59290600)

    I've worked as a staff scientist at a national lab for the last 30 years, and IMHO the work has gradually become less appealing. Its difficult to pin down a major cause - more budget tracking more regulations, etc, a general bureaucratization of the work. Attempts to squeeze $2B projects into $1B budgets that result in the work all feeling sort of slip-shod. Its even possible that budget squeezed projects are overall more efficient (though that is not at all clear to me), but they are less fun to work on.

    No one goes into astrophysics or similar fields for the money, they do it because the work is fun and interesting. As that declines, the field becomes less attractive relative to other fields I"m still here and not planning to leave, but I'd say the probability of my leaving has been gradually increasing with time. (though at the moment I have fun projects - though an insane level of work)

    The question of how much society should spend on fundamental research is of course a fair one.

    • After graduating with a master's degree in 2002, I spent a good part of the noughties looking for and trying out a number of PhD projects in physics. At a lot of places, it seemed I wasn't welcome -- they'd rather have one of the local "good guys" than me and my Cambridge degree. The ones I did start didn't work out for a number of reasons; the last one had a long commute to a noisy office where I'd do my actual work over ssh and X forwarding to a supercomputing centre. Of course, I had to be there every

  • "Yes, but in the end, you are predicting fashion trends. The same trends that are fickle, change all the time, and are possibly mocked by the next generation" You are describing "fast fashion" a way for garment companies to keep raking in the profits using planned (style) obsolecence. You won't have to wait a generation for the styles to get mocked; "Ewww Suzy Q! Those jeans are soooo last April!" Now if these brilliant minds can figure out a way to get the greedy fucks to stop using overseas fire trap f
  • by hoover ( 3292 ) on Thursday October 10, 2019 @06:06AM (#59291518)

    Nearly my entire physics graduation year (degree in 1996) ended up in IT. Higher pay, better job security and a better life / job balance made the decision a no-brainer at the time as companies (both corporate and start-ups) were hiring left, right and center at the time.

    I've kept astronomy as a hobby on the side. I decided in favour of IT when I discovered that the researchers at the uni were not working for a "common good" but the atmosphere was just like in any other company... so much for the myth of the ivory tower.

  • ...to tell your dad you're gay.

Every nonzero finite dimensional inner product space has an orthonormal basis. It makes sense, when you don't think about it.

Working...