Skip to content
Field notes

Notes from the work I am actually shipping.

What I am seeing in current builds, frontier AI, and software calls. Written so an operator can decide what to do next.

How we apply this

FeaturedAug 28, 20266 min read

A supply-chain-risk label is not a blank check to punish a vendor

Newsletter

Get new field notes in your inbox

Practical AI, shipping, and governance notes. No spam. Unsubscribe anytime.

Browse by focus

Primary hubs

The three piles I write in most. Counts are matching notes in the archive.

Full archive

Browse every field note

RSS

AI

Two Models Dropped Today. One Pattern Explains Both.

Google released Gemini 3.1 Flash-Lite and OpenAI pushed GPT-5.3 Instant on the same day. One is cheaper and faster than anything before it. The other is less annoying than anything before it. Together, they point at the same thing: AI in 2026 is getting better where it actually counts.

Mar 3, 2026 · 5 min read · Steve Defendre

AI

The Real AI News on March 2, 2026 Isn’t the $110B. It’s the Runtime.

OpenAI's February 27 funding blast is still the biggest AI story on Monday, March 2, 2026. But the number everyone keeps repeating is not the part I care about most. The Amazon deal, and the stateful runtime behind it, is the part that could actually change how AI gets used at work.

Mar 2, 2026 · 6 min read · Steve Defendre

AI

What OpenAI's $110B Round Actually Says About the AI Industry Right Now

OpenAI just closed $110 billion in new funding at a $730 billion valuation. Before you process the numbers, think about what it means for everyone building in AI — and everyone betting against it.

Mar 1, 2026 · 6 min read · Steve Defendre

AI

Anthropic vs. The Pentagon: When AI Safety Principles Meet Government Money

The Pentagon demanded Anthropic strip Claude's safety guardrails for autonomous weapons use. Anthropic said no. Now there's a lawsuit, a federal blacklist, and a question nobody in AI can avoid: what are your principles actually worth?

Feb 28, 2026 · 8 min read · Steve Defendre

AI

NVIDIA Just Posted $42.96B in Quarterly Profit. I Had to Read It Twice.

NVIDIA's latest results are absurd even by recent AI standards: about $42.96B in quarterly net income and about $120.07B for fiscal 2026, with revenue at $68.1B for the quarter and $215.9B for the year. This is what AI infrastructure demand looks like when it is no longer early.

Feb 27, 2026 · 6 min read · Steve Defendre

AI

Google's Nano Banana 2 Is Out — and It Changes the Math on AI Image Generation

Google launched Nano Banana 2 today, combining Imagen-quality output with Gemini Flash speed and real-time internet knowledge. For developers and creators, this is the first image model that actually knows what happened last week.

Feb 26, 2026 · 6 min read · Steve Defendre

AI

Perplexity Computer Is a Super Agent for Everyone Who Doesn't Want to Set One Up

Perplexity launched "Computer" yesterday — a cloud-based super agent that builds websites, writes reports, runs 19 AI models, and connects to 400+ apps with no local setup. This is what happens when agent-level AI gets consumer distribution.

Feb 26, 2026 · 7 min read · Steve Defendre

AI

Anthropic Just Launched an AI Security Scanner — and Cybersecurity Stocks Are Paying the Price

Anthropic's new AI-powered security scanning tool is shaking up the $200B cybersecurity industry. CrowdStrike, Palo Alto, and SentinelOne took hits as the market asks: does AI commoditize what these firms charge billions to do?

Feb 25, 2026 · 6 min read · Steve Defendre

AI

$650 Billion on AI in 2026: The Biggest Bet in Tech History Has a ROI Problem

Big Tech will pour $650B into AI this year. But 80% of companies see no productivity gains yet. What gives?

Feb 24, 2026 · 5 min read · Steve Defendre

AI

Stripe's Autonomous Coding Agents: 1,300 PRs a Week and the Future of Software Development

Stripe deploys AI agents generating over 1,300 pull requests weekly. Architecture, reliability strategies, and what this means for developers.

Feb 21, 2026 · 6 min read · Steve Defendre

AI

Google's Gemini 3.1 Pro: Revolutionizing Agentic AI and Complex Reasoning

Google DeepMind releases Gemini 3.1 Pro in preview, with 1M-token context, multimodal reasoning, and major gains for agentic workflows. Here is what matters for teams shipping real products.

Feb 20, 2026 · 6 min read · Steve Defendre

AI Agents

RentAHuman: AI Agents Are Now Hiring Humans for Real-World Tasks

RentAHuman, a new marketplace where AI agents hire humans for physical tasks, has exploded to 500k+ workers. This flips the automation narrative and bridges digital agents to the physical world.

Feb 19, 2026 · 5 min read · Steve Defendre

AI

OpenAI Frontier: Enterprise AI Agents Treated Like Employees

OpenAI launches Frontier, an end-to-end platform for building and managing AI agents in enterprises. Onboarding, access controls, and feedback loops mirror human employee management. Uber, HP already onboard.

Feb 18, 2026 · 6 min read · Steve Defendre

AI

NVIDIA's Physical AI: The ChatGPT Moment for Robotics Has Arrived

NVIDIA declares the 'ChatGPT moment' for robotics with new open models, Isaac GR00T enhancements, and industry-wide robot deployments. Practical takeaways for AI builders.

Feb 17, 2026 · 6 min read · Steve Defendre

AI Agents

Google's WebMCP: Ending Screenshot Hell for AI Web Agents

WebMCP turns websites into structured toolkits for AI agents. No more vision-based guesswork—direct JSON calls mean faster, reliable web automation. Here's what builders need to know.

Feb 16, 2026 · 6 min read · Steve Defendre

AI

DeepMind's Aletheia: Dawn of Autonomous AI Researchers

Aletheia signals a shift from chatbot outputs to full research workflows. Here is what matters for serious builders shipping in 2026.

Feb 13, 2026 · 8 min read · Steve Defendre