Ilya Sutskever – The age of scaling is over

Bullish:

• $DEFENSE - As AI becomes more visibly powerful, governments and the public will demand increased oversight and control, benefiting defense contractors and security companies

  • “I also maintain that as AI continues to become more powerful, more visibly powerful, there will also be a desire from governments and the public to do something.”

    Ilya·1:00:46
• $SEMIS - Despite the shift from pure scaling to research-focused AI development, massive compute requirements persist as the field requires 'big computers' for continued advancement

  • “Now from 2020 to 2025, it was the age of scaling or maybe plus minus, let's say, the arrow bars to those years, because people say this is amazing, you gotta scale more, keep scaling. The one word, scaling. But now the scale is so big... So it's back to the stage of research again, just with big computers.”

    Ilya·22:06
• $NVDA - Current AI models generalize dramatically worse than humans, implying continued massive demand for compute and hardware to overcome these fundamental limitations through scale

  • “So the thing which I think is the most fundamental is that these models somehow just generalize dramatically worse than people. Yes. And it's super obvious. That's that seems like a very fundamental thing.”

    Ilya·25:04
• $ROBOTICS - The massive gap between current robotic capabilities and human dexterity suggests significant investment and innovation will be needed in the robotics sector

  • “But to train a robot in the real world to quickly like pick up a new skill like a person does seems very out of reach.”

    Ilya·27:54

Bearish:

• $OPENAI - Current AI approaches will hit diminishing returns and 'peter out' as they cannot achieve AGI with existing methods

  • “I think what people are doing right now will go some distance and then peter out. It will continue to improve, but it will also not be it. So, the it, we don't know how to build.”

    Ilya·1:07:18
• $AISW - AI software companies risk stalling out with all offerings looking very similar across different companies, struggling to differentiate

  • “I think the rep I think it could I think it could stall out and I think stalling out will look like it will all look very similar Yeah. Among all the different companies, something like this.”

    Ilya·1:23:07