How to do research at the MIT AI Lab (1988) [pdf](people.cs.umass.edu)
people.cs.umass.edu
How to do research at the MIT AI Lab (1988) [pdf]
https://people.cs.umass.edu/~emery/misc/how-to.pdf
8 comments
Well this is an easy paper to summarize: Be good at everything , read everything, talk to everybody in the field, don't ever sleep, and be able to dilate time to fit everything in!
It's important to note that this paper was published over 25 years ago, the field of A.I. was smaller then. I've read this paper multiple times, and I don't think I ever got the sense that you would need to not sleep to do it all. They spend some time talking about how to determine which are the important things to do, which you can let slide (for instance, on page 3 they say "There’s a whole bunch of journals about AI, and you could spend all your time reading them. Fortunately, only a few are worth looking at." then go on to talk about which ones to look at).
Unfortunately this seems to be what a lot of people strive to do, then burn out or can never really produce impactful research. Sleep, rest, relaxation, are absolutely essential to creative work. You needn't be good at everything, but you will need to strive for the best in something, and find that thing nobody has ever done before.
I skimmed through it, and most of the advice sounds pretty solid. However, I suspect that following all this advice at once is not humanly possible (or at the very least, leaves you little time for anything else in your life). I did like the part about the 'cool people' secret paper club.
I find it interesting that this is published on a University of Massachusetts website rather than an MIT website.
I also think a hefty dose of this paper could apply to grad school in most fields at many top schools.
I also think a hefty dose of this paper could apply to grad school in most fields at many top schools.
It's been reposted lots of places. Here's an ostensibly more official source, if you prefer: http://dspace.mit.edu/handle/1721.1/41487