A breakthrough unfolds

Поделиться
HTML-код
  • Опубликовано: 11 сен 2024
  • In December 2019, DeepMind’s AI system, AlphaFold, solved a 50-year-old grand challenge in biology, known as the protein-folding problem. A headline in the journal Nature read, “It will change everything” and the President of the UK's Royal Society called it a “stunning advance [that arrived] decades before many in the field would have predicted”. In this episode, Hannah uncovers the inside story of AlphaFold from the people who made it happen and finds out how it could help transform the future of healthcare and medicine.
    For questions or feedback on the series, message us on Twitter @DeepMind or email podcast@deepmind.com.
    Interviewees: DeepMind’s Demis Hassabis, John Jumper, Kathryn Tunyasunakool and Sasha Brown; Charles Mowbray and Monique Wasuna of the Drugs for Neglected Diseases initiative (DNDi]) & John McGeehan of the Centre for Enzyme Innovation at the University of Portsmouth
    Credits
    Presenter: Hannah Fry
    Series Producer: Dan Hardoon
    Production support: Jill Achineku
    Sounds design: Emma Barnaby
    Music composition: Eleni Shaw
    Sound Engineer: Nigel Appleton
    Editor: David Prest
    Commissioned by DeepMind
    Thank you to everyone who made this season possible!
    Further reading:
    AlphaFold blog, DeepMind: deepmind.com/b...
    AlphaFold case study, DeepMind: deepmind.com/r...
    It will change everything, Nature: www.nature.com...
    AlphaFold Is The Most Important Achievement In AI-Ever, Forbes: www.forbes.com...
    Bacteria found to eat PET plastics, NewScientist: www.newscienti...
    Protein Structure Prediction Center: predictioncent...
    An interview with Professor John McGeehan, BBSRC: bbsrc.ukri.org...
    John McGeehan profile, University of Portsmouth: researchportal...
    Drugs for Neglected Diseases initiative (DNDi): dndi.org/
    A doctor’s dream, DNDi: • A doctor’s dream: A pi...
    The Curious Cases of Rutherford and Fry, BBC: www.bbc.co.uk/...
    Hannah Fry: hannahfry.co.uk/
    Find Seasons 1 & 2 on RUclips: dpmd.ai/3geDPmL
    Or search “DeepMind: The Podcast” and subscribe on your favourite podcast app:
    Apple Podcasts: dpmd.ai/2Rzlmcu
    Google Podcasts: dpmd.ai/3geDjp5
    Spotify: dpmd.ai/3w29cb4
    Pocket Casts: pca.st/30m1

Комментарии • 63

  • @platypao
    @platypao 2 года назад +36

    Wonderful! A video version of this would be appreciated :)

  • @haneulkim4902
    @haneulkim4902 2 года назад +9

    Seriously the best podcast I've ever listened to. Great info, well explained, good voice + music. Just perfect! Thank you×3000

  • @leighedwards
    @leighedwards 2 года назад +4

    I'm so impressed and grateful for the amazing work and contributions to science, and for the honesty of the DeepMind scientists about whether AlphaFold shows anything like true understanding or intelligence. More power to their elbows!

  • @zealeigh
    @zealeigh 2 года назад +4

    Its amazing to think Sci-fi concepts like nano bots already exist in biology as proteins with such simple building blocks. Thank you Hannah, Dan and everyone involved. Fantastic. Great to hear the AlphaGo team taking another big bite. I really enjoyed their documentary.

  • @ataraxia7439
    @ataraxia7439 2 года назад +2

    There is something kind of magical about seeing a huge intimidating problem just crumble before you as time passes.

  • @notmeprobably_
    @notmeprobably_ 2 года назад +9

    This AI solution is the most exciting one to me that Deepmind has come up with, and I didn't know if I could love anything more than Alphastar

  • @jimmyt_1988
    @jimmyt_1988 2 года назад +2

    This was incredibly inspiring. Thank you so so much guys. Amazing podcast.

  • @sparrow5686
    @sparrow5686 2 года назад +6

    Deepmind is so good at breakthrough's...

  • @barisbasar3909
    @barisbasar3909 2 года назад +1

    I love these podcasts. I really hope that one day I can contribute to solving the riddle of agi . Thank you for your effort !

  • @monolizm
    @monolizm 2 года назад +1

    just imagine a visualization to this material

  • @madi112233
    @madi112233 2 года назад +1

    Great progress, but IMHO we are still far from understanding the fundamentals of nature. Please keep up with great work!

  • @bhallos_baus
    @bhallos_baus 9 месяцев назад

    Saved to Watch Later

  • @youcaio
    @youcaio 3 месяца назад

    my humble so humble mind is hella in love with Hassabis Deep, so so deep one

  • @thehappycoder3760
    @thehappycoder3760 2 года назад +1

    I did enjoy that

  • @williamstrahan3198
    @williamstrahan3198 2 года назад

    Very interesting and Nice!

  • @max72122
    @max72122 2 года назад +1

    while listening to this podcast, It got me into thinking that will we be able to use ai to create a artificial human in the next few decades? I am not sure but I think it can be possible.

  • @cbaloch
    @cbaloch 11 месяцев назад

    Explained simply.

  • @palfers1
    @palfers1 2 года назад +2

    Hannah, don't be lazy. We need videos!

  • @chrism.1131
    @chrism.1131 Год назад

    Being a specialist in enough varied subjects eventually leads to becoming a generalist. AI >>> AGI.

  • @alejandrochataing5341
    @alejandrochataing5341 2 года назад

    Awesome podcast

  • @prsstnc
    @prsstnc 2 года назад

    Wooo Wooo!

  • @imranhossain3504
    @imranhossain3504 8 месяцев назад

    Good

  • @stonegarraty9315
    @stonegarraty9315 2 года назад +1

    I listen to these to fall asleep

  • @Pr0GgreSsuVe
    @Pr0GgreSsuVe 2 года назад +1

    Woo!

  • @DheerajYadav-tx7mk
    @DheerajYadav-tx7mk 2 года назад

    Videos are better than blank podcasts

  • @stonegarraty9315
    @stonegarraty9315 2 года назад +1

    Don’t you think it’s dangerous to give robots learning abilities?

    • @lokeshranjanjha1796
      @lokeshranjanjha1796 2 года назад

      A parallel AI safety would be required, along with excelling autonomous AI.

    • @ian4692
      @ian4692 2 года назад

      Do you think the fact that human beings can learn is dangerous?

    • @eeriemyxi
      @eeriemyxi 2 года назад

      @@ian4692 well, it is though

  • @Astrophynstein
    @Astrophynstein 2 года назад

    0:13 Samsung gang 😎

  • @zaewaznazari
    @zaewaznazari 9 месяцев назад

    ❤❤😅

  • @cleitondelimabarbosa70
    @cleitondelimabarbosa70 2 года назад

    Pqp tinha que ter em português rs

  • @Learner161
    @Learner161 Год назад

    Yes, please. Play 3D video

  • @jazzisback
    @jazzisback 2 года назад +3

    Fantastic news, but 92.4 is not enough. One could end up with an amyloid-like structure. As a physician, I wouldn't dare apply such a protein to a human's body. I hope the researchers can get to 100% accuracy. Nevertheless, an admirable success.

    • @zealeigh
      @zealeigh 2 года назад +3

      Even if it is not absolute results, I can only imagine how illuminating it is to all who are searching through the haystack for the needle. I do hope all involved have your mindset as well.

    • @isabellagoogleaccount8418
      @isabellagoogleaccount8418 2 года назад +2

      92.4 is literally more than many experimentally verified with x-ray crystallography lol get real.

    • @Phlarx
      @Phlarx 2 года назад +1

      The results of this process isn't the answer... it's a signpost pointing us toward likely answers. To expand on Rochow's analogy, this is drastically narrowing down the places in the haystack where the needle might be.

  • @stalinvlad
    @stalinvlad 2 года назад

    Prions. That's all I got

  • @neatodd
    @neatodd 2 года назад +4

    Protein folding and AI. Not simple concepts so why make it harder for the listener to follow what's being said by adding background music? Now if someone could invent an AI that removed background music and left the speech that would be a breakthrough.

    • @minhuang8848
      @minhuang8848 2 года назад +2

      You are aware this has been a thing since forever now, right?

    • @ataraxia7439
      @ataraxia7439 2 года назад

      I personally liked it. Maybe they can have two versions one with music one without?

    • @triton62674
      @triton62674 2 года назад

      I mostly didn't mind it but it was a bit much at times

    • @cconnors
      @cconnors 2 года назад

      Nvidia Broadcast has this feature if you have an RTX GPU.

  • @Mighty_Deeds
    @Mighty_Deeds 2 года назад +1

    Apparently Every animal & Single cell organisms have cracked this problem millions of years ago. Just ask the Dinosaurs 🦖 🦕

  • @DanielSMatthews
    @DanielSMatthews 2 года назад +3

    Here is a reality check for you with regard for how primitive AI and mechatronics still is. Show me a robot that can successfully change a nappy without befouling the entire room and or terrorising the patient or infant. That is a job done successfully every day by some of our lowest paid human workers.

    • @AboSir
      @AboSir 2 года назад +9

      Strange comment. Leave the child rearing to loving parents - not steel robots. Zoom out - AI has the potential to find cures to diseases that have ailed mankind for centuries. AlphaFold is also helping find a way to recycle plastic. Huge developments. Alas, there is always that one negative guy out there like yourself who dismisses everything without really even bothering to listen to the knowledge being dropped on the podcast. Don't be ignorant Daniel ;)

    • @DanielSMatthews
      @DanielSMatthews 2 года назад +1

      @@AboSir Your use of an ad hominem attack clearly demonstrates your ignorance and or lack of intellectual integrity because you had the opportunity to actually address the point I made which is entirely valid in that it does vividly and clearly illustrate the current state of the art's lack of sophistication and integration. Your argument is at best the equivalent of pointing at a box of parts and insisting that you should be proud of your ability to make a car from it, without ever having done so.

    • @vaprin2019
      @vaprin2019 2 года назад +2

      @@DanielSMatthews You're judging a fish by its ability to climb a tree. Even if we assume that no one could train an AI to successfully change a nappy, this doesn't mean that AI is "primitive". It just means that AI isn't very good for changing nappies. Anyone can come up with an arbitrary problem that is easy for X and hard for Y, and proclaim that Y is "primitive". I'm sure there are legitimate arguments for what you're trying to claim, but this isn't a good one.

    • @DanielSMatthews
      @DanielSMatthews 2 года назад +1

      @@vaprin2019 You have missed or deliberately avoided the point I made, twice now. It is the lack of sophistication _and_ integration that must be acknowledged if we are to act with intellectual integrity and not just be obnoxious fanboys. Look at how many computational units or columns are in the human brain and the complexity of the connectome required to integrate their functions. Each of these AI "breakthroughs" that tickle your novelty detector is not much more than a cluster of such units operating without the emergent synergy of the coherent whole we see in an actual brain. BTW Oxudercinae _can_ climb trees.

    • @Sleepy_Alligator
      @Sleepy_Alligator 2 года назад +1

      LOL! Just wow. Amazingly no added value, pointless comment.

  • @Dr.Z.Moravcik-inventor-of-AGI
    @Dr.Z.Moravcik-inventor-of-AGI 2 года назад

    Is it fantastic or fake-tastic? Let's make bets on deepmind and western world in general.