Jason A. Hoffman

Part 9 in a series. Previously: Model Eats the Software (Part 8), The Confident Incompetence Problem (Part 6), Confidence All the Way Down (Part 6b), On Keeping AI in the Critical Path (Part 7), The Disintermediation Principle (Part 5), Zen of Unix Tools (Part 4). The Transition Strategy has been narrative. Consultants build slide decks.…

February 24, 2026 Two recent pieces frame the AI infrastructure debate from opposite poles. Citadel Securities published “The 2026 Global Intelligence Crisis”, arguing that AI displacement risk is overstated and adoption follows predictable S-curves. Citrini Research published “The 2028 Global Intelligence Crisis”, modeling a structured bear case where rapid white-collar displacement triggers a deflationary spiral…

A Recursive Proof of the Confident Incompetence Problem Originally published at https://the-ai-dialogues.com/commentary/confidence-all-the-way-down I asked a frontier model to write the Constitution of the United States in formally verifiable code. The same kind of code we use to certify flight software — the kind that proves a helicopter autopilot won’t kill you. The model delivered 1,969…

When AI Architects Systems It Cannot Operate Originally published at https://the-ai-dialogues.com/commentary/the-confident-incompetence-problem I models have mastered a specific form of helpfulness: confidently proposing systems that require an Operations Research PhD to implement, delivered with the enthusiasm of someone who absolutely cannot help you implement it. Ask a frontier model to solve an optimization problem, and watch…

Part 8 in the series. Previously: [Zen of Unix Tools], [The Disintermediation Principle], [Keeping AI in the Critical Path]. In August 2011, Marc Andreessen wrote “Why Software Is Eating the World” in the Wall Street Journal. His thesis was that software companies were poised to take over large swathes of the economy: that the world…

What it means when the marginal cost of cross-referencing approaches zero. Every domain has a corpus too large for any human to hold at once. Law has 275,000 sections of statute and regulation. Medicine has millions of papers and trial results. Codebases have millions of lines across thousands of files. Financial systems have decades of…

Fine. It’s a bubble. I’ll give you that. The valuations are insane. OpenAI at $150 billion. Anthropic raising at $60 billion. Nvidia’s market cap swinging by hundreds of billions on a single earnings call. The infrastructure spend looks like the late ’90s fiber buildout. Journalists are writing “AI bubble” pieces with the same confidence they…

Harvard Business Review is running sponsored content titled “A Blueprint for Enterprise-Wide Agentic AI Transformation.” Accenture has committed $3 billion to its AI practice. McKinsey’s QuantumBlack has grown to 5,000 AI specialists. Wipro just launched a new operating model combining “advisory, AI, and enterprise transformation services.” Deloitte, BCG, Bain — every firm that sold you…

Or: I spent six weeks compensating for what Claude Code couldn’t do yet, and then it could. On Christmas Eve 2025 I started a repo called agentic. The first commit was 2,312 lines across 19 files: 14 agent roles — frontend engineer, security engineer, QA, product manager, technical writer — plus orchestration concepts, communication protocols,…