Paul Middlebrooks

@pgmid

Neuroscientist | Brain Inspired Podcast | Future alien

Nomad
Vrijeme pridruživanja: rujan 2014.

Tweetovi

Blokirali ste korisnika/cu @pgmid

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @pgmid

  1. 15. sij

    Wolfgang says the cup is almost empty. Which means lots of potential fill it. He advocates: - the separation of church and theory - the constraint of theory with neuroscience observations - the development of spiking neural networks. Amen!

    Poništi
  2. proslijedio/la je Tweet
    3. sij

    Thanks to for having me on Brain Inspired! We had a fun discussion around neuro/AI, the surprising emergence of human memorability behavior in deep nets trained to see, and what we think curiosity is good for (novelty-based RL).

    Prikaži ovu nit
    Poništi
  3. 3. sij

    I’ll never forget having Nicole on to discuss how brains and convolutional nets forget and remember images, how neuro and RL can team up to understand novelty and curiosity-driven learning, and waking up super early.

    Poništi
  4. 22. pro 2019.

    Do our cognitive limits promote our general intelligence? Tom Griffiths and his resource-rational framework have something to say about that. Plus bridging Marr’s levels, cognitive prosthetics, AGI, and how to greet people before fencing duels.

    Poništi
  5. 9. pro 2019.

    Close your eyes and imagine the art below... Thomas Naselaris knows what parts of your brain will get more oxygen, and he’s got the models to show how. Plus current state and promise of fMRI, AI and neuro epicycles, and careful dads at skate parks.

    Poništi
  6. 27. stu 2019.

    Glad is now part of my exposome... enjoyed learning about her handy RNNs, and how she uses them to model how our brains switch between behavioral regimes, and how we (and machines) multi-task. And now I know what an “exposome” is… maybe.

    Poništi
  7. 18. stu 2019.

    Jon Brennan breaks from reading Alice In Wonderland to talk about using linguistics, models, and EEG/fMRI to learn how brains process and understand language. Plus modern NLP, how linguistics can inform neuro and AI, and the goal to be less wrong.

    Poništi
  8. proslijedio/la je Tweet
    8. stu 2019.

    Listen to Kyle Dunovan, Ntrepid’s Computational Behavioral Scientist, talk about , , academic burnout, and other fascinating topics in “Academia to Industry,” a podcast from Brain Inspired:

    Poništi
  9. 6. stu 2019.

    Andrew eats deep networks for breakfast! Well, he chews their nonlinearities, spits 'em out, and watches what happens, then eats normal breakfast. We talk about what linear networks teach us about nonlinear ones. Plus replay, semantics, and patience.

    Poništi
  10. 28. lis 2019.

    Enjoyed talking with about using graph neural networks for construction, how people and machines might perform mental simulation, and how cognitive science isn’t a real thing (ahem, ha!) - plus what hat to wear when giving a talk.

    Poništi
  11. 18. lis 2019.

    I talk w about how cool the basal ganglia are although they get ~neglected in AI (minus dopamine/RL), his reflections on academia after transition to industry, the associated stigma. Closest I've come to crying on an episode, but I stood fast!

    Poništi
  12. 7. lis 2019.

    I talk with Phillip about his company Brainworks, which strives maximize neuroscience knowledge to build trustworthy AI, plus his early days at NASA (using neural nets in the 80s!), ethics, brain-machine interfaces, & the AI startups world

    Poništi
  13. proslijedio/la je Tweet
    27. ruj 2019.

    Happy thought of today; “If I have a religion my religion is science, I believe in science. If you look back over the human history, we have seen more progress in science than in any other domain of human culture” Liz Spelke on Braininspired

    Poništi
  14. 25. ruj 2019.

    Liz Spelke and I talk about core cognition in infants & how language might synthesize new abilities that make us humans think we're so special, and whether infant cognition could teach us how to make better AI. Just don't ask her how you should parent!

    Poništi
  15. 19. ruj 2019.

    Does anyone know who decided to play my podcast outro music at the closing mixer at ? That was surreal and awesome.

    Poništi
  16. 19. ruj 2019.

    Looking forward to talking again sometime!

    Poništi
  17. 13. ruj 2019.

    So weird to have “brain inspired podcast” on my name badge. Come say hi if you’re at and you see me around (or hear my voice I suppose)!

    Poništi
  18. 8. ruj 2019.

    I roped in to talk about how we transform physical sound waves into meaning in our minds, how on avg you speak 5 syllables/sec, speech vs. language, levels of understanding, and of course spot-on impersonations.

    Poništi
  19. 29. kol 2019.

    I talk with about building end-to-end robots using deep reinforcement learning at to do things like navigate cities. Also, challenges for embodied (robot) AI, the role of neuroscience, Vaucanson's duck, an iconic 80's film.

    Poništi

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.

    Možda bi vam se svidjelo i ovo:

    ·