Google’s DeepMind finds 2.2M new substances in materials science win - eviltoast

Trove of combos is >45 times larger than number unearthed in entire history of science.

  • Redkey@programming.dev
    link
    fedilink
    arrow-up
    12
    ·
    11 months ago

    Does anyone know if the AI also figured out how to make any of these materials, or if it only assessed that they could probably exist in a stable state? This article isn’t clear about that but it seems like the answer is the latter.

    • mkhoury@lemmy.ca
      link
      fedilink
      arrow-up
      2
      ·
      edit-2
      11 months ago

      It seems like they have an automatic lab that tested 58 of them and 41 were successfully synthesized. So 70% success

  • JohnDClay@sh.itjust.works
    link
    fedilink
    arrow-up
    9
    ·
    11 months ago

    I hope it indicates what they may be useful for? Because testing even a small percentage of those would be a huge undertaking.

  • AutoTL;DR@lemmings.worldB
    link
    fedilink
    English
    arrow-up
    6
    ·
    11 months ago

    This is the best summary I could come up with:


    The trove of theoretically stable but experimentally unrealized combinations identified using an AI tool known as GNoME is more than 45 times larger than the number of such substances unearthed in the history of science, according to a paper published in Nature on Wednesday.

    The number of substances found is equivalent to almost 800 years of previous experimentally acquired knowledge, DeepMind estimated, based on 28,000 stable materials being discovered during the past decade.

    Two potential applications of the new compounds include inventing versatile layered materials and developing neuromorphic computing, which uses chips to mirror the workings of the human brain, Cubuk said.

    The team deployed computation, historical data, and machine learning to guide an autonomous laboratory, known as the A-lab, to create 41 novel compounds from a target list of 58—a success rate of more than 70 percent.

    The key to the improvements was how AI techniques were combined with existing sources such as a large data set of past synthesis reactions, he added.

    The techniques outlined in the two Nature papers would enable new materials to be identified “with the speeds necessary to address the grand challenges of the world,” said Bilge Yildiz, a Massachusetts Institute of Technology professor who was not involved in either piece of research.


    The original article contains 592 words, the summary contains 209 words. Saved 65%. I’m a bot and I’m open source!

  • jeebus@programming.dev
    link
    fedilink
    English
    arrow-up
    4
    arrow-down
    1
    ·
    edit-2
    11 months ago

    And here I am with a Pixel 8 Pro and an unsuccessful transfer of data, apps, and payment methods from a Pixel 4a. My payment account got locked out and they want me to verify my identity yet again. I tried calling customer support but all it does is refer me to a fucking website. Fuck Google and their piece of shit hardware and customer experience.