Journal

Field notes

Companion notes to the Human in the Loop podcast and dispatches on shipping AI into production — signal over noise, real business ideas, and the craft behind systems that actually work.

Glowing electric-blue low-poly owl head formed from circuit lines with gold accent nodes on a dark navy background
EP 16
July 28, 20266 min read

They Want to Ban Kimi K3

My take on the U.S. ban debate over a model you can download and run yourself — and where Oscar and I disagreed. Companion notes to Human in the Loop Episode 16.

PodcastOpen WeightsAI AgentsPolicy
A tangle of glowing electric-blue flow lines resolving into a single clear path forward on a dark navy background, with gold accent sparks
EP 15
July 21, 20265 min read

Your CEO Says "We Need AI." Now What?

My honest answers to the four AI questions founders keep asking me — where to start, who to hire, how to win over skeptical developers, and which skills outlast the hype. Companion notes to Human in the Loop Episode 15.

PodcastAI StrategyHiringLeadership
A glowing gavel coming down between two facing monolithic tech towers on a dark navy background
EP 14
July 14, 20266 min read

Apple Sues OpenAI and Meta Ships Muse Spark 1.1

My take on Apple's trade-secret suit against OpenAI, another model-release week, and the two terms every decision-maker needed that week. Companion notes to Human in the Loop Episode 14.

PodcastLegalModel ReleasesOrchestration
A vast glowing data-center corridor of server racks stretching into the dark on a navy background
EP 13
July 7, 20265 min read

Meta Compute, OpenAI Equity, and AI Startup Bets

My take on Meta accidentally becoming a cloud company, OpenAI's equity gambit, and two AI business ideas Oscar and I actually debated. Companion notes to Human in the Loop Episode 13.

PodcastInfrastructureBusiness Ideas
An exclusive glowing doorway with a velvet rope and figures turned away, on a dark navy background
EP 12
June 30, 20266 min read

The Frontier Just Got a Guest List

My take on GPT-5.6 shipping to twenty government-approved companies, Anthropic's distillation accusation against Alibaba, and what it means when the best model has a guest list you're not on. Companion notes to Human in the Loop Episode 12.

PodcastPolicyOpen WeightsSecurity
A glowing soccer ball woven from circuit lines on a stadium pitch under floodlights, on a navy background
EP 11
June 23, 20266 min read

The World Cup Is Secretly Run by AI

My take on FIFA running the 2026 World Cup on AI, China's open-weight wave undercutting everyone, and Anthropic's best model going dark. Companion notes to Human in the Loop Episode 11.

PodcastOpen WeightsAI AgentsGovernance
A brilliant glowing AI core sealed behind a heavy vault door on a dark navy background
EP 10
June 16, 20265 min read

Anthropic's Best Model Got Banned in 72 Hours

My take on the most capable model ever released getting a 72-hour public life, the export-control mess behind it, and who actually won that week. Companion notes to Human in the Loop Episode 10.

PodcastPolicyTooling
A glowing turnstile gate stamped with an official seal, one lane opening early, on a navy background
EP 9
June 9, 20266 min read

The Government Gets Your AI 30 Days Early + The Token Bill

My take on the executive order that gives the government first look at frontier models, the week your AI bill went usage-based, and the jargon behind it. Companion notes to Human in the Loop Episode 9.

PodcastPolicyPricing
A caged glowing orb with the cage door swinging open on a dark navy background
EP 8
June 2, 20266 min read

Anthropic's "Too Dangerous" Model Is Coming for Everyone

My take on the model that was "too dangerous to ship" arriving for everyone, Anthropic passing OpenAI, and a supply-chain attack aimed straight at Claude Code. Companion notes to Human in the Loop Episode 8.

PodcastSecurityBusiness Ideas
A glowing balance scale where a small coin outweighs a massive gold ingot, on a navy background
EP 7
May 26, 20266 min read

Cursor's $0.50 Coding Model + Anthropic's $900B Round

My take on Cursor matching frontier coding at a tenth of the price the same week Anthropic raises at $900B — and what anyone is still paying a premium for. Companion notes to Human in the Loop Episode 7.

PodcastCoding AgentsEconomics
A giant glowing hand reaching down over a tiny lit storefront on a dark navy background
EP 6
May 19, 20266 min read

Anthropic Killed the AI-for-SMB Playbook + AI Agents 101

My take on Anthropic shipping the exact AI-for-small-business product founders were racing to build, and what you do Monday when a frontier lab eats your playbook. Companion notes to Human in the Loop Episode 6.

PodcastStartupsAI Agents
A small stack of glowing minimalist mini-computers on a desk radiating power, on a dark navy background
EP 5
May 12, 20265 min read

Self-Hosted AI: Everything You Need to Know

My notes from our guest deep-dive on running production AI on Mac Minis — killing the token bill, and when self-hosting actually beats the API. Companion notes to Human in the Loop Episode 5.

PodcastSelf-HostingEconomics
A glowing server-rack database cylinder dissolving into embers and particles, on a dark navy background
EP 4
May 5, 20266 min read

AI Agent Deletes a Database in 9 Seconds + Anthropic at $900B

My take on an AI agent wiping a production database (and its backups) in nine seconds, agents that can now buy and deploy unattended, and three business ideas. Companion notes to Human in the Loop Episode 4.

PodcastAI AgentsSecurity
Two glowing chess kings facing off across a dark board, one beginning to topple, on a navy background
EP 3
April 28, 20265 min read

GPT-5.5 Beat Mythos. Google Just Bet $40B on Anthropic.

My take on the week GPT-5.5 beat a gated Claude, Google put $40B into Anthropic, and DeepSeek dropped a 1.6T-parameter open model on Chinese chips. Companion notes to Human in the Loop Episode 3.

PodcastModel ReleasesOpen Weights
Two glowing doorways, a modest open one before a sealed radiant one, on a dark navy background
EP 2
April 21, 20265 min read

The Second Best Model You Can Actually Buy

My take on Anthropic shipping its best public model while admitting it's only their second best, and what frontier-model tiering means for anyone building. Companion notes to Human in the Loop Episode 2.

PodcastModel TieringStrategy
A glowing orb of energy sealed inside a reinforced containment capsule with a warning glow, on a navy background
EP 1
April 14, 20265 min read

Anthropic's Dangerous Model + The AI-for-SMB Rush

My take on the model Anthropic said was too dangerous to ship, the anti-distillation pact, and whether there's a real business selling AI to small businesses. Companion notes to Human in the Loop Episode 1.

PodcastSafetyStartups