Machines that Learn by Doing - Prof. Levine (Google Brain/Berkeley)

Поделиться
HTML-код
  • Опубликовано: 22 окт 2024

Комментарии • 21

  • @citiblocsMaster
    @citiblocsMaster 6 лет назад +28

    This guy could have been a rapper, he speaks so clean and fast that the automatic CC makes no mistakes. I'm glad he didn't tough because the content is even better.

    • @tahac805
      @tahac805 6 лет назад

      CC transcribes “soil tanker”

    • @citiblocsMaster
      @citiblocsMaster 6 лет назад +1

      the CC makes *almost* no mistakes

  • @TheHeavenman88
    @TheHeavenman88 6 лет назад +23

    Epic Ending !

  • @darrendwyer9973
    @darrendwyer9973 5 лет назад +2

    what I don't understand. the human brain works in analog and the neural network works in realtime over many many years. Why is AI research so determined to use trained neural networks, instead of a realtime feed of information over a long time.

  • @domsau2
    @domsau2 6 лет назад +6

    So, domestic robot in 5 years?

  • @ckhalifa_
    @ckhalifa_ 5 лет назад

    Fantastic talk! Clear and concise talk, very informative. And what an ending! Haha

  • @akompsupport
    @akompsupport 6 лет назад +1

    Ends on a high note! Awesome lecture! Is there a git repo of this work?

  • @ai-ur5uv
    @ai-ur5uv 4 года назад

    He is amazing

  • @willasn9080
    @willasn9080 6 лет назад +15

    only the intelligent people are watching this kinda of Videos and learning about this topic. I mean look at deepminds Channel it has only a few thousand views on most of their Videos. Arguably the most important and probably last really big invention is in it's birth Phase right now and most dont pay attention.
    nothing wrong watching cute panda Videos or some netflix documentary. but this is more important.

    • @VerisimilitudeDude
      @VerisimilitudeDude 6 лет назад +3

      Either intelligent people or bored and stoned people lol.

    • @willasn9080
      @willasn9080 6 лет назад +1

      Verisimilitude Dude
      or maybe bored, intelligent, stoned people.

    • @citiblocsMaster
      @citiblocsMaster 6 лет назад +1

      Do you think many people have read the bitcoin paper when it came out? Most important discoveries are completely neglected when they first come out, the recognition only comes in retrospect. The story of neural nets or LSTM is another good example (unpopular literally for decades).

    • @willasn9080
      @willasn9080 6 лет назад +1

      Benjamin Crouzier
      imagine a world, where deepminds channel would have over a million subscribers and where not basketball/soccer would be in the news, but who made the best research on Quantum Computing or the newest idea of new types of neural networks or the next breakthrough in terms of electric vehicles or CIRSPR .... beautiful! but i think in this world the average iq would be 120.

  • @machinistnick2859
    @machinistnick2859 3 года назад

    Thanks a lot😁

  • @AlinNemet
    @AlinNemet 6 лет назад +4

    the ending was lol

  • @irlserver42
    @irlserver42 6 лет назад +2

    🥇

  • @mymcoatesjourney8849
    @mymcoatesjourney8849 6 лет назад

    Just a comment on what I believe may be a disconnect between AI ALGORITHMS in GENERAL + how AI discerns & makes hiring decisions FOR HR Offices, in terms of "DISCRIMINATING" between FIT and UNFIT PEOPLE in the hiring process across the U.S via ONLINE EMPLOYMENT APPLICATIONS for the past 10 - 15 years. This based on resumes and potentially SS# or other sources. I am not sure what has been PROGRAMMED into Human Resource AI algorithms in terms = ETHICS & FAIRNESS HIRING for both the INTEL & MAC platforms. I just know what I have gone through For 10 YEARS+ is NOT NORMAL at ALL . . .
    I wrote a memoir. If you care to read more and want to get closer to the BOTTOM LINE, start at Chapter 23. It's on AMAZON & KINDLE = www.amazon.com/Tell-Me-THINK-GOING-ebook/dp/B07FK31R23/ref=sr_1_1?s=digital-text&ie=UTF8&qid=1532173141&sr=1-1&Keywords=Tell+Me+.+.+.what+do+you+think+is+going+on