10 ms·

And this is it. A reward system based on paper publishing and conference speeches is fundamentally broken. There are really no built-in checks and balances, u

• phkahler 6y ago
  >> A reward system based on paper publishing and conference speeches is fundamentally broken. After reading that I'm wondering, what is the reward?
  • konjin 6y ago
    You aren't fired - sorry you can't get fired from a tenure job - you don't get paid peanuts to fund research.
  • skosuri 6y ago
    Post-tenure? Peer-recognition & Funding.
  • joshvm 6y ago
    From a computer vision perspective. If you get a paper into CVPR, or SIGGRAPH, or NeurIPS, it has the potential to make your career. It's that important to some employers (academic and not) and that's why everyone struggles to publish in those "top" venues. It's not like you won't have a career without a paper in these places (I really want to stress this), but it's enough of a boost to your reputation that everyone wants to. Think of it like getting a Nature or Science paper, or something in Phys Rev, or Cell. It's about prestige. The real issue is conferences taking the place of academic journals. In most fields conferences are laid back places to share your WIP. In CS, conferences require full papers which means many ML researchers (for example) structure their whole year around meeting the NeurIPS and ICML/ICLR deadlines. It's a bit broken because those conferences have limited capacity and it can be a lottery if you get in. Journals don't have the same restriction and don't have deadlines unless you're going for a special issue, so you can actually take the time to polish before submitting.
    • breezest 6y ago
      I did not work in the hottest topics in Machine Learning (and not in computer vision) as I cannot find rooms for innovation. I did publish several papers in 2nd-tier conferences and submit papers to journals. Near the end of my graduation, I started to find jobs in the industry. I have been shocked when I saw some employers will target authors who published in these big conferences even for non-research positions. (It sounds like discrimination.) I always wonder if my research profile will meet their requirements.
      • snovv_crash 6y ago
        When I was hiring in industry for computer vision, by the end I honestly didn't even look at people's publications. GitHub, any skills they list, and examples of projects they worked on. What got published at the end seemed to have no impact on what was important for industry.
    • yodsanklai 6y ago
      > it has the potential to make your career. This is true, but I think for many researchers the main motivation is self-esteem. It's the way they are ranked among their peers. In other circles, it may be your bonus, or how fast you run a marathon. It may be seem silly but this is what drives many researchers.
  • asdff 6y ago
    A 6 figure job with excellent benefits that you can hold until you die, along with discount tuition for your family. Not a bad gig even with all the administrative bullshit professors are saddled with these days.
• kps 6y ago
  Peter Higgs, of the boson and Nobel: “It's difficult to imagine how I would ever have enough peace and quiet in the present sort of climate to do what I did in 1964. […] Today I wouldn't get an academic job. It's as simple as that. I don't think I would be regarded as productive enough.” — https://www.theguardian.com/science/2013/dec/06/peter-higgs-boson-academic-system https://www.theguardian.com/science/2013/dec/06/peter-higgs-...
  • sushshshsh 6y ago
    What a tragically sad inefficiency of our current society that people who are way smarter than me are toiling away in the dark for pennies an hour on something extremely hard and useful to society, only to have their results denied by the system. Meanwhile I'm being paid obscene amounts of money to center divs and type @Autowired and @Component for projects that never end up launching anyway. It's all so badly badly flawed.
    • throw9872349872 6y ago
      Thats why most of us are in industry ;)
      • sushshshsh 6y ago
        Money is an important motivator, but one wonders if it's possible to make amazing societal contributions and also make $2 million in a decade :)
        • hokumguru 6y ago
          I think those people end up in research work at large companies. Essentially doing academic work in the private sector.
          • sushshshsh 6y ago
            Now all that's left to do is democratize that and you have a perfect world, no? That is to say increase the pool of such jobs from 10,000 to 10,000,000
• skosuri 6y ago
  >>And this is it. A reward system based on paper publishing and conference speeches is fundamentally broken. What should it be based on? Should academics be judged on performance? If so, how?
  • raziel2701 6y ago
    I wish academics were challenged to bring a technology to life. A working gizmo cannot be argued away. We need some mechanism by which the products of research are exposed to reality, a little bit of a market force to help us evaluate what works and what doesn't. Already there is a replication crisis. What we need is to make talk less cheap, ideally give the academics skin in the game. This would be a system where people toil away for years before producing something, and some may never succeed, but when they do they get all the upside from creating a new technology, and we as a society benefit from having this new invention. Essentially, I would break apart the monolithic idea of academia. There's too many inefficiencies, we waste away the talent of very smart people working on incomprehensible details that only a handful of people care about, and they write papers no one will read, only to then go to finance or tech where they'll wonder what was the point of grad school since a small percentage of their skills/specialized knowledge is necessary to make a living. We're creating all these PhDs and we don't know what to do with them. We need to foster a more entrepreneurial path in these people, they're stuck thinking that they only have two choices: academia or industry.
    • adrianN 6y ago
      Bringing gizmos to life immediately disqualifies all fundamental theoretical research. And the humanities.
      • raziel2701 6y ago
        No, theoretical research is needed to understand why a gizmo works, we often use phenomenological models to progress, and the fundamental reason that ties everything together is figured out later. We didn't understand electricity and magnetism for a long time but we were creating all kinds of devices and materials via trial and error. The arrow of discovery has a strong record with experiment first, theory later. Metallurgy was pretty much figured out by trying shit out first, later came the formalization and explanations for why strain hardening is a thing, or doping and microstructure for example. For the humanities, yeah I have no idea. I didn't specify but I meant the science and engineering parts of academia. I think math doesn't have a reproducibility crisis since it should be all logical proofs that don't require infrastructure, so I guess math is fine as it is.
• baby 6y ago
  It’s funny because I relate this to the “too many meetings” of some parts of the private sector. The pressure comes from the challenge of finding time to do actual work.
• _wldu 6y ago
  Yes, it's basically a popularity contest these days.
• teorema 6y ago
  One basic problem is that science (and related disciplines) is really a slow incremental process where no one knows what is going on, and many people make small contributions, but this doesn't sell well. So every little thing ends up being promoted as if it is. It's the classic cargo cult problem: no one is actually a supergenius -- if there ever was such a thing really -- so we'll just act as everyone is because that's the corner we've painted ourselves into. Another problem is that academics is full of dead ends, and it has to be if it's being done right. But no one likes that either. So rather than working on changing expectations, we do the opposite and bean count. Yet another problem is the classic fan fiction / fan community issue: if you get into enough expertise, the experts have to become the audience, because no one else understands it, but then that leads to increasingly narrow fields of view because of increasingly narrow interests. You could broaden the audience, which is increasingly a demand, but that has downsides too, because often what's really the way forward is incomprehensible and boring to the average person. People love their iphones; not so much all the incremental computer science, physics, and engineering that went into every little part. Something I think this essay maybe misses is that there are a lot of senior academics who would say publishing doesn't matter, that a paper is a dime a dozen, and what does matter are grants, and lots of them. This is how this dynamic has shifted in a lot of places, from the science -> publishing -> money. You might argue that this is better but it has its own set of incentive problems.
  • Guthur 6y ago
    There is such an anti individualism these days that I find bonkers. It's as if people crave collectivism. There have been quite obviously individually remarkable people that stand head and shoulders above everyone else, and don't think that's changed. Nikola Tesla, Carl Friedrich Gauss, Isaac Newton, and many more. Feats of greatness may come from collaborating but there are often great leaders and minds at the centre of it.
• euske 6y ago
  But is it really possible to establish such a system that ensure correctness, innovation and public trust, and so on? It seems like to me that science is already an impossible endeavor from the beginning. This kind of thinking reminds me of scientific anarchism championed by Paul Feyerabend, like "Anything Goes". He strongly criticized the corporate thinking in the scientific community, but what else we can do? Maximizing the (ultimately) monetary benefit and its efficiency is the only way we can run any business, and everything else looks fishy to most people (including myself).
• N1H1L 6y ago
  What's even more broken is that impact factors and h-indices -- all that data analytics we use to measure science is fundamentally broken. They are worse than useless