What if the biggest economic divide created by AI is not between people who have access and people who do not, but between those who can afford the best intelligence and those who cannot?

Anthropic’s new economic model explores how AI could reshape jobs, wages and growth. But my own experience with a newer, far more expensive model raised a different question: if frontier AI becomes a premium resource, will it really democratize opportunity—or simply give the deepest pockets a bigger advantage.

The AI LEAD Act would do more than let consumers sue chatbot companies. It could change how AI software is designed, insured, purchased, and funded.

AI gives startups unprecedented engineering leverage, but if assurance becomes bureaucratic, incumbents regain the advantage. The real opportunity is to automate assurance so small companies can remain fast and still prove their software is safe.

We spent thousands of years building a civilization in which the human body is the ideal participant. Now we are building machines capable of participating in it. The real question is no longer whether robots can do our work, but what happens when millions of people can no longer earn a living because machines can do the same jobs faster, cheaper, and without needing a paycheck.

Don’t ask AI to understand the whole system every time. Architect the system so it doesn’t have to.

I have started to think of AI coding agents as extraordinarily capable engineering collaborators that have almost no natural fear of complexity.

AI can dramatically accelerate engineering, but speed only creates value when paired with context, human judgment, and verification. This is how I use AI as an engineering collaborator while remaining accountable for every system that reaches production.