Search
Buy your pass
Keynote
AI for Good Global Summit

(Replay) Closing the 100,000 year “data gap” in robotics

  • Register

    * Register (or log in) to the Neural Network to add this session to your agenda or watch the replay

    • Date
    23 January 2026
    Timeframe
    16:00 - 16:20 CET
    Duration
    20 minutes
    • Days
      Hours
      Min
      Sec

    This is a replay of the session that took place on the Frontier Stage during the AI for Good Global Summit in Geneva, Switzerland from 8 to 11 July, 2025. AI is rapidly advancing the way we think, but we live in a material world. We still need to move things, make things, and maintain things. We need AI-driven humanoid robots to support an aging human population that does not have enough workers.  Large models based on internet-scale data can now pass the Turing Test for intelligence. In this sense, data has “solved” language and many analogously claim that data has solved speech recognition and computer vision.

    Will data also solve robotics?  Rich Sutton points out in the “Bitter Lesson” that data has surpassed all the best-laid analytic work in AI, and I accept that this trend will eventually hold sway in robotics. But when?

    Using commonly accepted metrics for converting word and image tokens into time, the amount of internet-scale data used to train contemporary large vision language models (VLMs) is on the order of 100,000 years. However, the data needed to train robots is not yet available.  One way to obtain such data is through human teleoperation, but the largest such dataset reported is on the order of 1 year. This suggests that humanoid robots will be available in…100,000 years.

    This 100,000-year “data gap” could be closed in three other ways: simulation, spatial analysis of internet videos, and collecting data from real robots operating in real environments.  Simulation works well for robots that fly or walk, but is notoriously unreliable for robot manipulation.  Spatial analysis of videos is progressing, but currently far too noisy to be useful for robot manipulation.  The last option – where data is collected as real robots operate in real commercial environments — requires bootstrapping with AI and “good old-fashioned engineering” to create robots with real return on investment that will be adopted by industry.  Such robots can create a “data flywheel” to increase performance and enable new functionality, accelerating the timeline to achieve reliable, general-purpose robots.

    Share this session with your network
    • Session starts in
      Days
      Hours
      Min
      Sec
    Add to Calendar
    Discover more from programme stream
    AI for Good Global Summit
    (Replay) Soft robotics for good: From marine species to ocean...

    (Replay) Soft robotics for good: From marine species to ocean...

    16 January 2026 - 16:00 to 16:20

    (Replay) Closing the 100,000 year “data gap” in robotics

    (Replay) Closing the 100,000 year “data gap” in robotics

    23 January 2026 - 16:00 to 16:20

    (Replay) Mind over machine: When AI listens to your thoughts

    (Replay) Mind over machine: When AI listens to your thoughts

    30 January 2026 - 16:00 to 16:20

    (Replay) Restoring voice and representation through inclusive AI

    (Replay) Restoring voice and representation through inclusive AI

    6 February 2026 - 16:00 to 16:20

    (Replay) Foundation models for autonomous driving

    (Replay) Foundation models for autonomous driving

    13 February 2026 - 16:00 to 16:20

    Related stories

    Are you sure you want to remove this speaker?