• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
  • HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    AI
    Reaction

    An AI-inspired case for studying the brain at scale

    Replying to a post suggesting some systems neuroscientists feel “research depression,” a commenter urges larger efforts to understand the brain’s model of the world and self.

    DG
    DT
    2 Sources, ,

    TLDR

    One poster says conversations with colleagues have reinforced a feeling that many algorithm-focused systems neuroscientists face a “research depression.” In a reply, another commenter argues that AI’s impact came from scaling transformers to solve problems. They hope larger-scale neuroscience could help explain intelligence and consciousness, and see foundation models as useful for combining data across sessions and subjects—not as an answer to understanding the brain.

    Combined views

    2.9K

    2 Sources, first seen 7h ago

    Combined views

    2.9K

    2 Sources, first seen 7h ago

    75 likes
    7h ago
    first seen 7h ago
    75 likes
    2 comments
    18 saves
    11 reposts
    2 comments
    18 saves
    11 reposts
    Featured Source

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet

    2 Sources

    @doristsaoIf someone had simply published an academic paper on transformers, no one would have cared. The impact has come from scaling it to solve actual problems. In neuroscience we similarly tackle problems at such a small scale (describing the neural correlate of percept X or task Y) that it's hard to make people care. It's hard to even know if any proposed algorithm is right. But I feel so inspired by what's happened in AI...if we also scale our efforts and aim for understanding the *full structure of the brain's model of world and self*, there actually is hope to solve the entire problem in our lifetime...and when we do, we will not only figure out intelligence (as LLMs have), but consciousness. What is the brain's algorithm for tracking a persistent piece of matter? What is the brain's algorithm creating programs with variable arguments? What is the brain's data structure for representing "I"? I am dying to know. Re the neural foundation model program: I think intelligently stitching data together across sessions and subjects is important, and foundation models are a good tool for that. But I absolutely disagree that they are an answer to understanding the brain.7h
    @dileeplearningRT @doristsao: @ShahabBakht If someone had simply published an academic paper on transformers, no one would have cared. The impact has come…3h

    2 Sources

    @doristsaoIf someone had simply published an academic paper on transformers, no one would have cared. The impact has come from scaling it to solve actual problems. In neuroscience we similarly tackle problems at such a small scale (describing the neural correlate of percept X or task Y) that it's hard to make people care. It's hard to even know if any proposed algorithm is right. But I feel so inspired by what's happened in AI...if we also scale our efforts and aim for understanding the *full structure of the brain's model of world and self*, there actually is hope to solve the entire problem in our lifetime...and when we do, we will not only figure out intelligence (as LLMs have), but consciousness. What is the brain's algorithm for tracking a persistent piece of matter? What is the brain's algorithm creating programs with variable arguments? What is the brain's data structure for representing "I"? I am dying to know. Re the neural foundation model program: I think intelligently stitching data together across sessions and subjects is important, and foundation models are a good tool for that. But I absolutely disagree that they are an answer to understanding the brain.7h
    @dileeplearningRT @doristsao: @ShahabBakht If someone had simply published an academic paper on transformers, no one would have cared. The impact has come…3h