How To Do Research At The MIT AI Lab | Reflections and Advice on Getting Into AI Research

Ғылым және технология

In this somewhat meta video, we're looking at a guide on how to do AI research at MIT from 1988, and discussing how my journey into AI research has aligned or diverged from this guide thirty years later. Plus, some things I wish I’d known when I started!
Sign up for CuriosityStream and Nebula here: curiositystream.com/jordan
Twitter - / jordanbharrod
Facebook - / jordanbharrod
Instagram - / jordanbharrod
Become a Patron! / jordanharrod
Thank you to Kasia, Jeff, Gerald, Milan, Ian, Becky, Jino, Daniel, Narskogr, Jason, and Mariano for being $5+/month Patrons!
Sources:
How To Do Research At The MIT AI Lab - dspace.mit.edu/bitstream/hand...
PhD in Pictures (Dent in the circle comic) - matt.might.net/articles/phd-sc...

Пікірлер: 34

  • @JordanHarrod
    @JordanHarrod4 жыл бұрын

    It's still finals week over here, so I thought I'd do a slightly more relaxed video on an interesting guide that I recently discovered. How has your field changed in the last 30 years?

  • @adelmomorrison3517
    @adelmomorrison35173 жыл бұрын

    Awesome, thanks so much for the information. I have been looking for something like this everywhere!

  • @jasong3972
    @jasong39722 жыл бұрын

    This is exactly what I was looking for, thank you so much!

  • @duncanhaywood4624
    @duncanhaywood46242 жыл бұрын

    Thank you for the great advice.

  • @guimaraesalysson
    @guimaraesalysson3 ай бұрын

    Great video

  • @polarbear986
    @polarbear9862 жыл бұрын

    It was very helpful!

  • @Moinbusiness
    @Moinbusiness3 жыл бұрын

    At last got a channel that talks so much about AI.Thanks

  • @zhubajie6940
    @zhubajie69404 жыл бұрын

    I've found that to truly know something you have to be able to explain it both in writing and in speaking to others. I remember my TA days and thesis writing. The discipline forces you to understand it. One of my best teachers I had was a Nuclear Reactor teacher who taught me one on one as a special topics class as few people wanted to learn it having 5 for the first semester and myself for the second part. So I read the material and explained it to him. A tough course but I learned the material very well.

  • @JordanHarrod

    @JordanHarrod

    4 жыл бұрын

    Definitely agree - I was a TA as an undergrad, and I really didn't fully understand the material until I had to teach it. That's one of the reasons why I like teaching (and why I do KZread)!

  • @rujanshakya9777
    @rujanshakya97774 жыл бұрын

    I have done bachelor in IT and I am always interested in AI fields and research, thanks for your advice and suggestions that are really helpful.

  • @nimbleline
    @nimbleline4 жыл бұрын

    I think this is all good general advice; I'm in industry and it all still applies!

  • @JordanHarrod

    @JordanHarrod

    4 жыл бұрын

    Good to hear! Especially as someone who doesn't plan to stay in academia :)

  • @phumlanimbabela-thesocialc3285
    @phumlanimbabela-thesocialc32852 жыл бұрын

    Thank you very much.

  • @JordanHarrod
    @JordanHarrod4 жыл бұрын

    Quick Clarifying Comment - In academia, you're considered in your "Early Career" in the 5-10 years after getting your PhD. PhD students are considered trainees. So, I've been doing some combination of self-taught/formally educated AI research for 3-4 years, but in the context of what is considered Early Career, I would consider myself new to the field, especially since I started in applied machine learning and moved into theory in the last year or so. There are definitely people who have been doing machine learning work for less time than me who would argue otherwise, especially if you're in industry, but I think that characterization of my progress with respect to the larger field is accurate within academia.

  • @aizan0

    @aizan0

    3 жыл бұрын

    So it means anyone become a.i researcher without having any education background

  • @rashedulkabir6227

    @rashedulkabir6227

    Жыл бұрын

    You are doing AI research or machine learning research?

  • @user-wr4yl7tx3w
    @user-wr4yl7tx3w Жыл бұрын

    More videos like this would be interesting

  • @mostechroom9780
    @mostechroom97803 жыл бұрын

    Doing Research In Deep Learning (South Africa) - Thank you for the advice

  • @rashedulkabir6227

    @rashedulkabir6227

    Жыл бұрын

    What is the difference between AI and deep learning?

  • @mostechroom9780

    @mostechroom9780

    Жыл бұрын

    @@rashedulkabir6227 AI covers the entire branch and deep learning is specifically geared towards neural networks

  • @rashedulkabir6227

    @rashedulkabir6227

    Жыл бұрын

    @@mostechroom9780 If I study AI will it discuss about deep learning?

  • @user-vn7ce5ig1z
    @user-vn7ce5ig1z4 жыл бұрын

    I got interested in A.I. over 20 years ago and was told I'd need to study computer-science and psychology. My interest in A.I. waned over the years, but I liked psychology, so I kept it and finished both. Considering what A.I. is like these days, I'm not sure how useful psychology is for it, maybe a little, but not a whole degree's worth. 🤔 🤷

  • @jacksparrow8417

    @jacksparrow8417

    2 жыл бұрын

    Did you study A.I and psy anymore

  • @riteshpanditi3635

    @riteshpanditi3635

    2 жыл бұрын

    I think it's more relevant that you'd believe. In the pursuit of building Artificial general intelligence, the understanding of the human mind is more important than even. Answering questions like what's consciousness ( agreed, more neuroscience related), how humans function and why we take the decisions and a lot more has yet to be figured out before we ever start to even comprehend what AGI is, the progress of AI needs psychology, period. PS: currently pursuing a Master's in Applied Artificial Intelligence

  • @adnanbakather4384
    @adnanbakather43843 жыл бұрын

    Thank you. Excellent brief. I am wondering if you can explain areas of research in ML? In other words, what are mostly asked questions in ML studies? It is about algorithm math, model evaluation, performance, feature engineering, use case, etc.

  • @percyblakeney3743
    @percyblakeney37433 жыл бұрын

    Two sense: To help AI receive a better understanding of awakening within our conscious experience have it read The Velveteen Rabbit by Margery Williams. It may also help those working in both research and development as well.

  • @barisakdogan2274
    @barisakdogan2274 Жыл бұрын

    As a MIT graduate student, what books would you recommend for AI learning ?

  • @flaryos9325
    @flaryos93252 ай бұрын

    Hello! I’m gonna do a research on AI, what advice do you can give me and what steps I should take?

  • @anngu2533
    @anngu25332 ай бұрын

    Thank you so much Jordan! It’s so valuable. I’m working on a paper in AI and healthcare, and I would love to hear your opinion. Can I send it to you?❤

  • @connorkapooh2002
    @connorkapooh20023 ай бұрын

    hey, i know this video is old so no worries if you dont see this. ive been trying to do my own independent research but im not sure how to build a research agenda or experiments or anything like that, that is taking a high level goal and turning it into something actionable. do you have any tips or advice around this? all the best

  • @MuhammadKhalid-jw1gd
    @MuhammadKhalid-jw1gd Жыл бұрын

    how to generate research idea from articles?

  • @abhishekghaskata9130
    @abhishekghaskata91304 жыл бұрын

    If you have your slack group, then share the link please

  • @JordanHarrod

    @JordanHarrod

    4 жыл бұрын

    I was referring to my lab/department slack groups - sorry for the confusion!

  • @abhishekghaskata9130

    @abhishekghaskata9130

    4 жыл бұрын

    @@JordanHarrod ok no problem