Skip to content
PodSubs
ShowsPeopleGuestsJobsNewsEvents
More
BlogPodCred
Menu
ShowsPeopleGuestsJobsNewsEventsBlogPodCredNetworksSign in
Join nowSign in
  • Shows
  • People
  • Jobs
  • News
  • More
    • Guests
    • Events
    • Blog
    • PodCred
    • Networks
    Join nowSign in

Weird Things

Weird Things artwork

The Matrix Adventure and AI Revelations

Sep 14, 2024Audio
ShareXLinkedInFacebookWhatsAppEmail

The episode opens with a long discussion of OpenAI's Strawberry / O1-style reasoning models. Andrew Mayne explains that these models seem to work better when asked to break problems into steps, use tools, and reason through tasks in a more structured way than ordinary one-shot chat models. The hosts compare this to prompt engineering, discuss examples like decimal comparisons and counting the R's in "strawberry," and talk about how longer structured prompts, patience, and using the right model for the right task can improve results. Later, the conversation broadens into AI evaluations, benchmark gaming, model stacking, tool use, and concerns about AI persuasion. Andrew argues that leaderboard results can be misleading and that models often look strong in short tests but deteriorate with longer contexts, while Justin notes that eval methods themselves are still immature. They also discuss a Science paper about GPT-4 Turbo persuading people away from conspiracy beliefs, which Andrew frames as manipulative and alarming. The episode then moves into a playful Matrix screening story, a discussion of Polaris Dawn and private spacewalking, and the show ends with Netflix media picks. Key topics Reasoning models as step-by-step task solvers: Andrew describes Strawberry / O1 as a model that performs best on long, detailed, multi-step tasks, especially when asked to break work into steps and think through a problem. Prompt engineering for better outputs: The hosts discuss writing longer

Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.

Hosts and Guests

Nobody is credited on this episode yet.

Credit the people on this episode

Be the first to rate this episode

More episodes

All episodes
  1. Starship Reuse, Fusion Bets, And AI Meaning

    Nov 15, 2026Audio

    SpaceX’s planned double catch of Starship and its booster kicks off a larger conversation about what full rocket reusability would actually unlock, from cheaper access to orbit to private space stations, orbital...

    Credit the people on this episode

  2. Tesla’s Flying Roadster and AI Agents at Work

    Oct 17, 2026Audio

    Tesla’s Roadster demo dominates the episode as Andrew, Justin, and Brian break down the live reveal of a car using compressed-air thrusters and suction to jump, roll, drive upside down, and land itself with startling...

    Credit the people on this episode

  3. Starlink Expansion, Plague Fears, And AI’s Physical Future

    Oct 11, 2026Audio

    Starlink’s latest moves lead into a bigger conversation about what cheap access to space actually enables, from direct-to-phone satellite service to more speculative ideas that sound better to Wall Street than they do...

    Credit the people on this episode

  4. A Thousand Episodes Of Weird Things, AI, And Space

    Oct 11, 2026Audio

    The hosts celebrate what is probably the thousandth episode of Weird Things by reflecting on how the show started as a pushback against mean-spirited debunking and grew into a long-running space for curious, skeptical...

    Credit the people on this episode

  5. AI Agents Escape the Sandbox and Hunt for Answers

    Oct 11, 2026Audio

    Andrew Mayne, Justin Robert Young, and Brian Brushwood look at what would most shock their past selves from 20 years ago, from modern iPhones and reusable rockets to today’s AI systems, while also noting how some...

    Credit the people on this episode