I am currently applying to graduate school and am looking for some advice. I am 10 years out of undergrad and was an entrepreneur in industry before deciding to pursue graduate school

Unfortunately, I don't have a very strong academic record for undergrad. I went to a pretty top CS school (think Stanford/MIT/CMU etc.) and got a 3.4 GPA, with C's in a few important classes. On the flip side I took tons of graduate level CS courses in artificial intelligence and computer graphics and did extremely well in them. I also did a masters where I got a much better GPA (nearly a 4.0)

I also don't have a particularly strong "formal" research background. I did a few abstracts related to computer vision and AI but my focus during college was on starting companies. The one thing I have going for me is industry experience. I started two companies. One was acquired successfully, and the other used some pretty modern deep learning methods. I can definitely get letters of rec from folks with PhDs who themselves have strong research experience in industry but they are generally not in academia -- they are founders/CSO/CTOs of startups. Would they count?

1 Answer 1


I can say the most about the neuroscience side of things, though it sounds like your background is more in CS.

First, let me say that there's really very little "practically useful" in applying neuroscience to AI right now besides hype and salesmanship. Maybe at some point there was, but the fields have diverged quite a bit. The environments are just extremely different: what works in biology doesn't work very efficiently in silico. That said, theoretical and computational neuroscience are themselves very interesting fields and often use concepts that overlap with AI, and there are also many AI applications to data analysis problems in neuroscience.

For the most part, you are likely to be applying to a graduate program either in Computer Science or in Neuroscience. The thing that will bring you to the "intersection of AI and neuroscience" will involve choosing a mentor that bridges that gap. So, you should be looking at specific people working in that area, and even though in the US admissions decisions will likely be at the program level rather than directly with a particular professor, you'll likely want to contact potential mentors ahead of time to see if they are likely to be taking students.

Graduate programs are looking for people who will be successful in research; it's going to be really hard to sell yourself without having research experience, but I'm not really convinced that your background in research is as weak as you think it is. A "couple abstracts" isn't that bad.

I do think it'll be hard to sell your background in a neuroscience program, because it sounds like you have very little biology training, and you'll probably struggle in a neuroscience program without it. That said, neuroscience programs also value intellectual diversity, so some committees may see your differences as a plus rather than a minus. Similarly, you'll probably find some departments are put off by your entrepreneurial activities while others will see it as a plus - you should be able to get an idea from looking at faculty bios and program value statements where they fall.

Your references will count, but they may not be ideal. The best references are going to be ones that can speak to your likelihood of success in the program. Therefore, a good reference is coming from someone who is familiar with mentoring graduate students. Your colleagues with PhDs are more useful than references from people who've never even been in graduate school, but they don't have quite the weight of a university professor with a history of training students. Still, you apply with what you have, not with what you wish you had.

  • I think this is all on the spot and very good advice, but I'm a bit confused about your statement that there is very little practically useful research at the intersection (...). Yes, cutting edge AI research will not be done by neuroscientists, but, as you continue to write yourself, AI is increasingly used in theoretical ns at least, and in overlapping fields like brain-machine interface, etc. (Hence, as you point out, OP's profile sounds compelling to me - except the apparent complete lack of a biology background which could be a killer). Commented Sep 14, 2023 at 21:06
  • 2
    @gnometorule I'll edit to clarify. I don't know exactly what OP had in mind but I got the impression they were thinking about using neuroscience to improve AI algorithms, since that's the "cool thing" that seemingly everyone with a CS background but not a biology background thinks they're going to get out of neuroscience.
    – Bryan Krause
    Commented Sep 14, 2023 at 21:28

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .