Just a casual question regarding MIRI
Currently I am planning to start a mathematics degree when I enter university, however my interest has shifted largely to computational neuroscience and related fields, so I’m now planning to switch to an AI degree when I go to study. Having said that, MIRI has always posed interesting problems to me, and I have entertained the thought of trying to do some work for MIRI before. And so my question boils down to this: Would there be any problem with taking the AI degree if I ever wanted to try my hand at doing some math for MIRI? Is a maths degree essential or would an AI degree with a good grasp on mathematics related to MIRI work just as well? Any thoughts or musings would be appreciated :)
Working on MIRI’s current technical agenda mostly requires a background in computer science with an unusually strong focus on logic: see details here. That said, the scope of MIRI’s research program should be expanding over time. E.g. see Patrick’s recent proposal to model goal stability challenges in a machine learning system, which would require more typical AI knowledge than has usually been the case for MIRI’s work so far.
MIRI’s research isn’t really what a mathematician would typically think of as “math research” — it’s more like theory-heavy computer science research with an unusually significant math/logic component, as is the case with a few other areas of computer science research, e.g. program analysis.
Also see the “Our recommended path for becoming a MIRI research fellow” section on our research fellow job posting.
If you check out their recommended knowledge base then you can see how much of it overlaps with your possible courses. For anything more specific, you could contact them directly; they often say how happy they are to discuss how to study for their technical agenda.
Study computer science. Math is instrumental and something you will pick up as you go along. All you need for that is some books, a notebook, and a pencil. Operational hands-on experience with artificial intelligence is something that you can get more out of from your studies in university. It will also open more doors should you choose to go a different route at a later point in time.
Not got good advice on the general question, but very much disagree with:
You need to self motivate yourself strongly to pick up new mathematical fields to the level where you can contribute. I’d make sure you had math courses in formal logic, whatever else you had; that would make it easier to learn the relevant stuff later. It also helps with thinking about these issues.
I studied computer science and I wish I had studied math; I think I’d find computer science much easier to “pick up as I go along”.
Seconding this, most people don’t just pick up math along the way. Maybe “math” in the “recipe book” sense.
I appreciate the great feedback from all of you, thank you :) I do have another quick question, but it’s of a lower priority. As of right now, I currently hold no degree. I’ve always been kind of Interested in the MIRI workshops, but I’ve always been nervous about signing up to one because: 1. I’m not sure if a degree would be necessary to keep up with the level of work people are to be involved in at the workshop and 2. In case my first point turned out to be true, I certainly wouldn’t want a student who had no real formal (I’ve still learning computability and logic and have started to branch out into set theory and similarly related fields) experience in the kind of Math MIRI deals with to be a nuisance to people trying to get some work done by asking them questions all the time. So here is my question stated in full ‘Would I be allowed to participate in a MIRI workshop, given that I have no degree as of right now, and could this factor be to the detriment of others there?’ Again, a lower priority question, but any comments or thoughts from users would be welcomed graciously :)