Genie 4, Token Metrics and AI Drug Research

Stacked Podcast28m 55s
Nick Saraev and Jack Roberts discuss AI token metrics and world models.
0 comments · 0 votesOpen discussionClose discussion
Sign in to join the discussion

    Nick Saraev and Jack Roberts discuss Google’s Genie 4 demonstration of interactive generated environmentsA world model learns how an environment is structured and how it may change in response to actions., including language steering and audio. Their proposed uses in entertainment and simulation are possibilities rather than demonstrated capabilities.

    They discuss a reported Amazon memo warning against measuring AI adoption by token usage. The hosts argue for considering work quality and time savedEvaluation is the systematic assessment of a system's behaviour or results against explicit criteria. while allowing room for experimentation, and use Goodhart’s law to explain how targets can distort a metric.

    They discuss Anthropic-reported preclinical drug researchAI-assisted scientific discovery uses AI to support hypothesis generation, experiment design, analysis, simulation, literature work, and interpretation while researchers retain responsibility. and possible future laboratory automation. The work described is early, and the hosts acknowledge that the reported results have not yet been tested in people; speculation about longevity is not evidence of a clinical benefit.

    Original YouTube thumbnailWatch on YouTube

    Video topic coverage

    • Agents & workflows0
    • Coding & tools0
    • Models & learning86.9
    • Infrastructure & compute0
    • Research & evaluation100
    • Safety & security0
    • Business & productivity88.1
    • Creative & media70.2
    • Robotics & perception21.4
    • Society & policy0

    Share this video

    Nick Saraev and Jack Roberts discuss AI token metrics and world models. With WWW.ARTIFICIAL-INTELLIGENCE.VIDEO, 28 October 2026 and duration 28m 55s.

    Nick Saraev and Jack Roberts discuss interactive world models, token-use incentives and early drug research.