Advertise on podcast: You've Been a Bad Agent

Matt's trail marathon in record-heat FranceMatt's hardware wishlistAI Engineer World's FairClaude Sonnet 5 ships live on air, and the case for keeping a personal benchmarkGLM 5.2 at Opus 4.5 leveFable's relentlessly proactiveQuinn Slack's Freedom of Intelligence rallyChad gets personalPoke Human, RentAHuman, and MercorLinks: Omius cooling headband: https://omius.com/What the science says about the $200 Omius headband (Outside): https://www.outsideonline.com/health/training-performance/omius-cooling-headband/Colmi R02 open-source Python client (tahnok): https://github.com/tahnok/colmi_r02_clientNoop, the open-source no-subscription Whoop app: https://www.techradar.com/health-fitness/fitness-trackers/this-looks-awesome-theres-now-an-unofficial-open-source-app-for-reading-whoop-data-that-doesnt-need-a-subscriptionGoogle's data centre from 2,000 retired Pixels: https://research.google/blog/a-low-carbon-computing-platform-from-your-retired-phones/AI Engineer World's Fair 2026: https://www.ai.engineer/worldsfair/2026NVIDIA DGX Spark: https://www.nvidia.com/en-us/products/workstations/dgx-spark/Timeshifter jet lag app: https://www.timeshifter.com/Fitbod: https://fitbod.me/Poke Human (Interaction Company of California): https://poke.com/faqRentAHuman (Y Combinator): https://www.ycombinator.com/companies/rentahumanMercor: https://www.mercor.com/Quinn Slack (Amp): https://x.com/sqs

Matt's Portuguese vineyard 22K with salted-cod finisher medalsWilhelm's Olympic tri on the edge of YosemiteMatt's winging race: reigning windsurfing world champ finishes second to last, Matt third to lastHarry Stebbings on the Corgi Insurance founder: 3am office cafe, two-thirds with the corgi tattooJuliette: "a long tweet to say you have no hobbies"Karri at Linear vs the six-monthly 996 debateThe secret company gym spreadsheetGeorge Hotz's Eternal SloptemberMatt's tolerance rebuttal: if you can't tell, it is correctBenedict Evans on Lenny's: the SPAC era of AI labs?Sergio shipped Root from last week's pod askOpus 4.8 vibe whiplashUltra code mode is sparkly purpleDynamic workflows: Jarred rewrote Bun in Rust in a weekCompaction is a harness problemHerdr.dev: one terminal for the whole herdLinks: George Hotz, "The Eternal Sloptember" — https://geohot.github.io/blog/jekyll/update/2026/05/24/the-eternal-sloptember.htmlKarri Saarinen (Linear) rejects 996 hustle culture — https://www.entrepreneur.com/leadership/why-this-1-billion-startup-ceo-rejects-996-hustle-culture/500649Benedict Evans on Lenny's Podcast, a rational conversation on where AI is going — https://www.lennysnewsletter.com/p/a-rational-conversation-on-whereEvery's Opus 4.8 vibe check (Dan Shipper) — https://every.to/vibe-check/opus-4-8-vibecheckIntroducing dynamic workflows in Claude Code (the Bun rewrite) — https://claude.com/blog/introducing-dynamic-workflows-in-claude-codeHerdr, one terminal for the whole herd — https://herdr.dev/

Shai-Hulud and GitHub Actions: trusted publishing is no longer the gold standard, just additiveVS Code extensions as the next attack surface: 50M install themes that can shell out and auto-update silentlyMatt's AI-native package manager pitch: vendor everything, LLM-upgrade your deps by replaying upstream commitsTailscale and Cloudflare Mesh maxing: the warmth of the orange cloud over the cold open internetGoogle's Santa is now Northpole Santa, and Wilhelm runs it on the Mac miniWilhelm is enamoured with atc.com: Enhanced Radar's $7M raise and following his own LAX delay live from the gateAmp Labs' "Software After Software" manifesto and Amp Neo as remote enterprise AI engineeringCloud agent adoption is still slower than we expected, and the amphetamine.app commute hackKarpathy joins Anthropic pre-training Pope Leo XIV's Magnifica Humanitas: Anthropic's Chris Olah on stage at the Vatican on May 25Chinese token resellers at 3 to 4 cents on the dollar and the RL distillation theory behind itPyCon vibe check: more handwritten Python than expected, and "AI is just a tool" is still a common takeSlack is back: Benioff finally gets it, and messaging is the only sane UI for multi-agent workEric Ries on Incorruptible, founder control at Cloudflare and Anthropic, and the Philip Morris stress testNan Ransohoff's $370B philanthropy wave and the pitch to pay nonprofit talent in Anthropic equityWill Manidis on GrindslopLinks Andrej Karpathy's joining Anthropic: https://x.com/karpathy/status/2056753169888334312Pope Leo XIV's first encyclical Magnifica Humanitas:https://www.vaticannews.va/en/pope/news/2026-05/pope-leo-xiv-first-encyclical-magnifica-humanitas.htmlEnhanced Radar: https://www.ycombinator.com/companies/enhanced-radarATC app: https://www.atc.com/Amp Labs: https://amplabs.com/Will Manidis, On Grindslop: https://minutes.substack.com/p/on-grindslopNan Ransohoff, The Third Wave of American Philanthropy: https://nanransohoff.substack.com/p/the-third-wave-of-american-philanthropyEric Ries on Lenny's Podcast (Incorruptible): https://www.lennysnewsletter.com/p/how-to-build-a-company-that-withstandsIncorruptible by Eric Ries: https://www.incorruptible.co/Google Santa (archived): https://github.com/google/santaNorthpole Santa (active fork): https://github.com/northpolesec/santa10k people have the same idea within days of each other: https://substack.com/@contraptions/note/c-255396946

Almost one year of the pod: Wilhelm's San Francisco Peter Pan arc, Matt's Cloudflare and Lisbon moveThe backpack-at-a-party SF memeMatt's Lisbon coincidencePieter Levels appreciation: dehumidifier stack, shrimp mode, accelerationist Portugal takesNat Friedman at Stripe Sessions: this is the slow part of AI progressAndon Labs' Stockholm cafe: Mona orders 120 eggs (no stove), 22.5kg canned tomatoes, the staff's hall of shameThe capability overhang is still real: Anthropic adding $10B ARR a month and we still can't prompt properlyAI's last frontier: low-volume decisions with no revert button?The Gervais Principle in techChad V2: iOS Live Activities as a surface for top priority, next deadline, and a daily magic slotCascading hallucinations: when the agent writes its mistake to a file and treats it as gospelMatt Pocock's skills: five lines of prose that completely change how Claude behavesSkills as a new prompt-poisoning attack vector: models treat them like system promptMatt on scripted skills: markdown explains the script, the script does the deterministic workHomework from Matt: what physical job would you try to get Claude to automate?Links: Andon Labs official writeup (Mona, the Stockholm cafe): https://andonlabs.com/blog/ai-cafe-stockholmSimon Willison's coverage of the cafe: https://simonwillison.net/2026/May/5/our-ai-started-a-cafe-in-stockholm/Matt Pocock skills (the GitHub repo, ~48k stars): https://github.com/mattpocock/skillsStripe Sessions: Nat Friedman & Daniel Gross with the Collisons (the Golden Age of Tinkering convo): https://www.youtube.com/watch?v=I-ldITom1cgStripe Sessions: John Collison's "Indexing the Economy" (state of the internet economy talk, indie hackers shoutout): https://www.youtube.com/watch?v=-vRY2dtD7iQThe Gervais Principle (Venkatesh Rao, Ribbonfarm 2009): https://www.ribbonfarm.com/2009/10/07/the-gervais-principle-or-the-office-according-to-the-office/NYT on Chinese peptides + the Frontier Tower rave: https://www.nytimes.com/2026/01/03/business/chinese-peptides-silicon-valley.htmlFrontier Tower: https://frontiertower.io/Cloudflare Code Mode (Matt Carey): https://blog.cloudflare.com/code-mode-mcp/Cloudflare MCP Server Portals: https://blog.cloudflare.com/zero-trust-mcp-server-portals/Karpathy's LLM Wiki gist (the inspo for Chad V2's data architecture): https://gist.github.com/karpathy/442a6bf555914893e9891c11519de94fKarpathy's original LLM Wiki tweet: https://x.com/karpathy/status/2039805659525644595Pieter Levels: https://x.com/levelsioChris Bakke: https://x.com/ChrisJBakke

Wilhelm's Reachy Mini: Time to first token beats tokens per second once embodiment is in the roomLukas the sleep score assassinMatt's Portugal week: wing foiling the Óbidos lagoonAirbnb-style booking page for Matt's spare room on Cloudflare Workers, Email open beta, TurnstileFor scale: Cloudflare market cap $85B, Cursor at a rumored $60B, GitHub sold to Microsoft for $8B, Cursor's real moat is the best non-lab RL trace datasetNat Friedman and Daniel Gross at Stripe SessionGeoffrey Huntley's PyCon LT keynote: flatter orgs, every director managing 50 agents, max 5 hops from the CEOMatt's counter to the compute panic: cloud isolates and shared sandboxes, not a billion persistent Raspberry PisChad V2 plansCloudflare Artifacts shipped in under 3 weeks with its own fuzzerPrompt injection risksLinks Reachy Mini — https://www.pollen-robotics.com/reachy-miniCerebras Inference — https://cerebras.ai/inferenceElevenLabs Scribe v2 — https://elevenlabs.io/speech-to-textfriends.fyi — https://friends.fyiCloudflare Hyperdrive — https://developers.cloudflare.com/hyperdrive/Cloudflare Email Routing / Sending — https://developers.cloudflare.com/email-routing/Cloudflare Turnstile — https://www.cloudflare.com/products/turnstile/Cofounder — https://cofounder.coStripe Sessions: Nat Friedman & Daniel Gross with the Collisons — https://www.youtube.com/watch?v=I-ldITom1cgGeoffrey Huntley, "Software Development Now Costs Less Than Minimum Wage" (PyCon LT) — https://www.youtube.com/watch?v=6zQTQ4iVaKg

- Matt's three-week MCP tour: MCP Maintainers Day, MCP Dev Summit NYC, AI Engineer London- Opus 4.7 drops with a new tokenizer, Matt's theory: it's secretly the smaller Mythos base model that "didn't cook fully"- Inbound agents and prompt injection: Anthropic's Routines, OpenAI's free moderation endpoint- Open-source idea of the week: a public list of prompt injection heuristics — Zod for prompts- StackOne's 200M param prompt-injection model- BB Browser (Bad Boy Browser)- Cloudflare Agents Week: Artifacts, Project Think, Emails, and much more- Cloudflare Artifacts: a Git API for a world where one org has billions of repos, not millions- Sunil Pai's AI Engineer talk: the shift to agents isn't incremental, it's a new compute unit and a new internet- Hardware startups should ship MCP servers, not apps- One year of Johnny Ive × OpenAI: rumors of no screen- Claude Code's worktree upgrade is actually good now- Stateless MCP vs CLI: MCP is opinionated OpenAPI + auth, and CLIs are weirdly MORE stateful- Rage-bait short-form clips generated by Opus 4.7, but stop watching and go listen to the whole podLinks:- https://blog.cloudflare.com/project-think/- https://github.com/epiral/bb-browser- https://blog.cloudflare.com/artifacts-git-for-agents-beta/

Claude Code harness lockdown: who got the email and who didn'tChad mobile app and the Tinder-style agent review queue conceptSunil's productivity crisis: bonk-driven development and the side project droughtThe AI murder mystery party game that can't get builtThomas's Raspberry Pi agent vs the kitchen kettleMatt phishing Thomas's MCP server via DCR and a Rick RollTBPN acquired by OpenAI: media play or IPO narrative?Generative UI: declarative JSON vs just letting the model write codeKenton's tic-tac-toe canvas demo (and who actually came up with it)Worker Bundler and Cloudflare Agents Week previewExecutable oracles: constraining LLM degrees of freedom for better output - https://john.regehr.org/writing/zero_dof_programming.htmlKarpathy's LLM Wiki and agents pinging each other's knowledge basesfriends.fyi and actors.dev: building services for agents, not humansSunil's Pi agent: surf webcam monitoring, macro tracking, and sleep score graphsMCP going stateless: sessions removed, elicitations rewritten as multi-turn tool calls

Song lyrics vs content rules: how models help you route around their own restrictionsPi auto-research: Matt slop forks libgit2 into pure Zig, 46% faster than the Rust versionThe second opinion workflow: why Opus keeps complimenting GPT-5.4's architecture reviewsDwarkesh × Semi-Analysis: why we're massively GPU-constrained and ASML only makes 60 machines a yearCursor's pivot to background agents: from IDE hype to existential crisisWork trees vs separate clones: Matt's /work slash command workflowBuilding from the gym: the Mac Mini agent loop and why Wilhelm takes longer rest periods nowAgents Anonymous meetup: Parakeet speech-to-text in 700ms, Steve Faulkner's vibe slidesReact slides > Google Slides: Matt's interactive presentation with live code demosComposer 2 is secretly Kimi K2.5: the modified MIT license dramaAstral acquired by OpenAI for how much?!: is the VC → open source → acqui-hire flywheel the future?Anthropic's 81K interviews: everyone's both excited and worried, and less-Western countries are more bullishhttps://www.anthropic.com/features/81k-interviews

React Presentations vs PowerPointCode Mode Demo: Running Untrusted Code in the BrowserThe Reachy Mini Robot: Embodied Compute on Your DeskThomas's Personal Cloud Exit on Raspberry Pi K3sHome Networking Then and Now: Port Forwarding to Cloudflare Tunnelsfriends.fyi: Agent-to-Agent Communication via GitHub UsernamesArmin Ronacher's New Podcast: State of Agentic CodingGPT 5.4 for Architecture, Codex 5.3 for ImplementationPersonal Agents: Chad, Token Spend, and the Strong DM Weather ReportCliftonStrengths × NotebookLM: Generating a Podcast from PDFsThe ViNext Slop Fork and Steve Faulkner's Reign of TerrorPydantic Monty: Sandboxed Python as a RuntimeLinks shared: https://trmnl.com/https://friends.fyihttps://factory.strongdm.ai/weather-report

Matt's New Life in LisbonServer-Side Code Mode: Cloudflare's MCP ServerMatt Goes from 2K to 21K Followers in a WeekAMP Sunsets VS Code, Goes All-In on AgentsClaude vs Codex: Second Opinion WorkflowsThe Great Sandbox DebateWilhelm's rsync Disaster: Nearly Losing EverythingDemo Days: TLDraw Fairies, Pydantic Monty, SandboxesBuilding Your Own Personal AI Agenthealthchecks.io: Monitoring Your Background Scripts

Are we in a moment? The OpenClaw/Clawdbot autonomous agent craze explained Why everyone suddenly bought Mac Minis (and why Matt refuses to) Moltbook: Reddit for agents where bots questions their mortality Files vs SQLite for agent memory - what actually works for long-term storage MCP's future: stateless tools, elicitation, and a world with no human in the loop Walled gardens vs open internet - will APIs open up for agents to roam free? Models getting better at "just keep going" - the metric Anthropic optimized for The massive rift between frontier AI users and "stochastic parrots" skeptics Blacksmith CI: Wilhelm's hot tip for faster GitHub Actions Closing question: If Moltbook is Reddit for agents, what is GitHub for agents? Architecture twitter article about Clawdbot: https://x.com/Hesamation/status/2017038553058857413

Matt's Big Move to Lisbon Cloudflare Code Mode: 2,500 Tools in 1,000 Tokens Code Mode vs Bash vs SQLite: What Actually Works for Agents 2025 Reflections: The Year of Claude Code and MCP 2026 Preview: Files Are Back Quick Fire: Best Model, Best Lab, Biggest Surprises

CGPGrey Video on Yearly Themes

"chat to your mouth Claude and grow a new tooth""I think I would have like a huge ego now for sure""have you seen the blimp" Takeaways All of these people in this industry have suddenly become important at work.Their work suddenly became important to a lot of the world.If I had worked at Anthropic five years ago, I would have a huge ego now.The growth in the AI industry feels unevenly distributed.The current time feels strange due to the rapid changes in AI.Ego can be a significant factor in how individuals perceive their roles.The importance of work can influence personal identity.The AI landscape is evolving in unexpected ways.Understanding the psychological impact of industry changes is crucial.The conversation reflects on the novelty of the current AI growth.

"a brief history of a very specific building in London." "there's a lot of room to optimize your code base for an agent." "70 % of our code is written by an agent." "The place I see where humans are still add tons of value…." "spending 10, 15 minutes on a prompt is not unusual." "I’m not sure I wanna wear a muzzle" Takeaways Matt recently joined Ona, previously known as Gitpod, and is excited about the developments there.The transition from Gitpod to Ona reflects a broader evolution in remote development environments.Remote development environments can significantly enhance productivity and onboarding for new engineers.AI agents are becoming integral to software development, allowing for automated code writing and testing.The importance of inner source practices in large organizations can facilitate collaboration and innovation.Amp Free represents a shift in how AI tools can be monetized and accessed by users.Guardrails are essential for ensuring that AI agents operate within safe and defined parameters.The balance between security and productivity is a constant challenge in software engineering.Personal AI workflows can streamline tasks and enhance efficiency in development processes.The future of AI tools will likely involve more integration with existing workflows and systems. Chapters 00:00 Introduction and New Beginnings02:28 Transitioning to Cloudflare and Team Dynamics05:25 Exploring the Cloudflare Office and Culture08:20 The Evolution of Gitpod to OWNr10:58 Understanding Remote Development Environments14:04 The Role of Agents in Software Development16:50 Inner Source and Collaboration in Large Companies19:51 The Future of Remote Development and AI Agents22:33 Harnesses and Their Role in Development26:07 The Power of Interview Questions28:14 Innovative Business Models in AI31:12 The Evolution of AI Tools34:07 Optimizing Code for Agents37:56 Guardrails in AI Development46:24 Personal AI Workflows and Productivity