'a lot of computer systems research is solving problems that only huge tech companies have (eg. fast datacenter networking, big data analytics systems), rather than working on technologies that empower individual users and the underprivileged. Frankly, the big tech companies don't need help from academics. … It would be much better if academics prioritised research that benefits those who *don't* have the resources to do this work themselves.'

classy yet

@22 Problem is, research is funded by people interested in those Big Problems. In order for researchers to research for non-big there has to be funding available.

@nomad Dr Narayanan, who Martin is quote-tooting, talks about how this happens even without explicit funding—e.g., students want to get jobs at BigCos so they gravitate to that work even if nobody's paying them to.

@22 ok.. figure out a way to put the money in the hands of people who you deem need solutions most. Unless of course you want to solve *that* problem too.

@Shitlord Dr Narayanan, in the thread that Martin was quoting, talks about other ways this happens even without money changing hands—so even if the "more needy" had money to fund research (at which point they wouldn't be needy), this dynamic might not change: Martin is super-classy for being conscious of this and choosing to go against the grain, and for scolding other researchers who have this choice (i.e., nobody's funding them) but fail to make it.

@22 This is so true. This is why I write compilers to help funding-poor weather and climate scientists accelerate their simulations.

Unfortunately the current narrative for funding is "economic impact". But that need not stop us. It is still possible to do the research that matters.

@22 For a long time I have been mulling over an article I want to write, something like "The politics of compiler writing", to discuss exactly this point: that even fundamental CS research is not value-free, because we choose who it benefits.

@22 And it is not just individual users and the underprivileged: it is also users with insufficient technical skills. Taking that into account from the start matters too.

@wim_v12e this is so awesome, thank you for doing so much good!

This note about inclusivity is so great. In my little way, I try to include as many as possible also, when I publish a project that might be interesting to non-coders (like Japanese stuff etc.) my instructions include how to install Git and Node or etc.

@22 Well, doing good is the aim, but I am not quite there yet ^_^
A problem with academic research is that there is no incentive to take software to production level. So it usually stops at the proof-of-concept phase.

@22 My personal current problem is that I have no time at all to work on compilers, because I am effectively a manager. Luckily this will be over in 6 months, and then I have great hopes for research!

Sign in to participate in the conversation

The social network of the future: No ads, no corporate surveillance, ethical design, and decentralization! Own your data with Mastodon!