エピソード

  • 260: Reinventing the Wiki
    2026/05/23

    What's the point of a personal wiki when AI agents have already read the entire internet? This week is about patterns for co-maintaining a personal wiki with an AI agent, and where its job ends and yours begins.

    Follow the show and be sure to join the discussion on Discord! Our website is workingcode.dev and we're @workingcode.dev on Bluesky. New episodes drop weekly on Thursday.

    And, if you're feeling the love, support us on Patreon.

    Mentioned in this episode:

    • Karpathy's LLM Wiki gist

    With audio editing and engineering by ZCross Media.

    Full show notes and transcript here.

    続きを読む 一部表示
    1 時間
  • 259: Walling Off Your Mental Garden
    2026/05/14

    This week is a philosophical conversation about what you let into your head — the podcasts, the influencers, the political figures, the friends — and the difference between a snapshot you can hold at arm's length and a living relationship you're quietly being shaped by.

    Follow the show and be sure to join the discussion on Discord! Our website is workingcode.dev and we're @workingcode.dev on Bluesky. New episodes drop weekly on Thursday.

    And, if you're feeling the love, support us on Patreon.

    Mentioned in this episode:

    • Fritz Haber (YouTube)

    With audio editing and engineering by ZCross Media.

    Full show notes and transcript here.

    続きを読む 一部表示
    1 時間 6 分
  • 258: On the AI of It All
    2026/05/07

    Adam wrote a blog post this week about why AI keeps making him feel worse. This week is an honest conversation about the hypocrite's bind: needing to use the thing you recognize the harm of, and what that does to you.

    Follow the show and be sure to join the discussion on Discord! Our website is workingcode.dev and we're @workingcode.dev on Bluesky. New episodes drop weekly on Thursday.

    And, if you're feeling the love, support us on Patreon.

    Mentioned in this episode:

    • Adam's blog post: "On the AI of It All"
    • Impeccable Style
    • Mental Notes

    With audio editing and engineering by ZCross Media.

    Full show notes and transcript here.

    続きを読む 一部表示
    57 分
  • 257: We Solved Everything
    2026/04/30

    "We have solved all of the world's problems." Observability past the point where more logs stop helping, continuous deployment when the customer is the federal government and the change-management board is a real room with real people in it, JSON's loose schema as a load-bearing feature rather than a quirk to apologise for, and the awkward question of who actually owns the code you wrote on a work-issued machine.

    Follow the show and be sure to join the discussion on Discord! Our website is workingcode.dev and we're @workingcode.dev on Bluesky. New episodes drop weekly on Thursday.

    And, if you're feeling the love, support us on Patreon.

    With audio editing and engineering by ZCross Media.

    Full show notes and transcript here.

    続きを読む 一部表示
    1 時間 9 分
  • 256: The Productivity Trap: How Optimizing Everything Broke Us
    2026/04/23

    It's 11:30 PM and you're still prompting. The code is better than anything you'd have shipped a year ago. The output is prolific, the roadmap is moving, the tooling is miraculous. So why are you still here at 11:30 PM? This week is an honest conversation about the strange shape of burnout in the AI era — the productivity that keeps compounding, the bar that keeps rising right along with it, and the quiet feeling that somewhere in the last year the craft stopped feeling quite like yours.

    Follow the show and be sure to join the discussion on Discord! Our website is workingcode.dev and we're @workingcode.dev on Bluesky. New episodes drop weekly on Thursday.

    And, if you're feeling the love, support us on Patreon.

    With audio editing and engineering by ZCross Media.

    Full show notes and transcript here.

    続きを読む 一部表示
    47 分
  • 255: AI in the Streets
    2026/04/09

    You spend all day steering AI through code. Then you step outside and it's steering everything else. AI is listening to therapy sessions and suggesting treatments to your therapist. Your spouse is arguing with a chatbot about where Savannah, Georgia is. You call a company for help and get handed from one AI pretending to be human to another AI pretending to be human. The crew has been noticing it everywhere, and this week they compare notes on what it actually feels like when AI stops being a tool you chose and starts being a thing that just happens to you.

    Links mentioned in the show:

    • hairstyles.net
    • RunPee.com

    Follow the show and be sure to join the discussion on Discord! Our website is workingcode.dev and we're @workingcode.dev on Bluesky. New episodes drop weekly on Thursday.

    And, if you're feeling the love, support us on Patreon.

    With audio editing and engineering by ZCross Media.

    Full show notes and transcript here.

    続きを読む 一部表示
    1 時間 1 分
  • 254: Claudependent
    2026/04/02

    The productivity gains are real. So is the nagging feeling that something else might be happening. The crew use AI every day, and this week they sit with a question they can't quite shake: when the tool handles more and more of the thinking, what does that do to the person using it?

    Follow the show and be sure to join the discussion on Discord! Our website is workingcode.dev and we're @workingcode.dev on Bluesky. New episodes drop weekly on Thursday.

    And, if you're feeling the love, support us on Patreon.

    With audio editing and engineering by ZCross Media.

    Full show notes and transcript here.

    続きを読む 一部表示
    53 分
  • 253: The Adversarial Agents Are Arguing Again
    2026/03/26

    What if the best way to get good work out of AI is to stop being nice to it? Adam and Tim have both landed on the same uncomfortable discovery: when you pit AI agents against each other, with fake points, opposing incentives, and competing models, the output gets dramatically better than anything a single polite prompt can produce.

    Adam's bug-hunting pipeline hands fake rewards to sycophantic agents and then throws the scores in the trash. Tim made Claude and ChatGPT argue for twelve rounds straight until they both said "ship it".

    Follow the show and be sure to join the discussion on Discord! Our website is workingcode.dev and we're @workingcode.dev on Bluesky. New episodes drop weekly on Thursday.

    And, if you're feeling the love, support us on Patreon.

    With audio editing and engineering by ZCross Media.

    Full show notes and transcript here.

    続きを読む 一部表示
    1 時間 6 分