More than three years after ChatGPT's release, only 27% of executives say AI has met their ROI expectations. The history of factory electrification explains why — most companies are at the light-bulb stage, adding Copilot licenses rather than reconceptualizing their businesses around AI. In this episode I map the three stages of AI adoption, and show what it actually takes to move from chatbots to the autonomous company — the only stage where the moat becomes real.

Published in early March 2026, Andrej Karpathy's autoresearch AI tool makes autonomous scientific experimentation cheap and easy — but it was designed to solve machine learning problems. I wanted to see if I could apply its loop architecture to my own work: refining my worldview, testing arguments, solving business problems.  In this video, I share how I adapted Karpathy’s autoresearch loops for problems that aren't easy to quantify, how to avoid the local minima trap, and the broader impact of these kinds of methods. I covered:

Last week Jensen Huang shared the numbers from NVIDIA’s order book: AI compute demand has grown a millionfold in two years. Much GCT coverage focussed on chips, robots, data centers in space, but I think Jensen revealed something far more important in his keynote: “the inference inflection has arrived,” and this is about to transform how all companies should manage their budgets. The inference era is already the operating assumption of the world’s most valuable company.

Apple may have stumbled into one of the most defensible positions in AI. This was not on my radar – just two months ago, I was describing a credibility crisis at the company; they appeared wrong-footed on the most important technology of our times and an acquisition was their only plausible way out. In this episode I work through what I and many other commentators missed – and what road lies ahead for Apple.

AI has become so embedded in how I work that I can no longer cleanly separate it from my thinking. That raises a question I find genuinely unsettling: is intensive AI use making me a sharper thinker, or quietly doing the opposite? In this episode I pull back the curtain on my full research and writing process — the custom tools, the friction points, and the places where I'm still not sure I've got it right. For Ezra Klein, having AI summarize material is a disaster for original thought. But my AI systems are designed to protect the cognitive work that has to stay human, while they handle everything else. Knowing where to draw that line turns out to be the hardest and most important question.

Meet R Mini Arnold - my OpenClaw chief of staff, which manages the equivalent of a ten-person team from a Mac mini in my garden studio. While I slept, that AI team debugged its own code at 3am, researched a trending Substack essay using five parallel investigators, and wrote a 4,600-word script for this very episode in 40 minutes. The gap between people who've started building this way and those who haven't is widening every week.

This is the first episode of AI Vistas, a new series where I bring together people I trust and respect to tackle a major question collectively. Today’s question: are we in charge of our AI tools, or are they in charge of us? Joining me are Nita Farahany, distinguished professor of law and philosophy at Duke University and a leading thinker on cognitive liberty and mental privacy; Eric Topol, founder of the Scripps Research Translational Institute and one of the world's most cited medical researchers; and Rohit Krishnan, engineer, former hedge fund manager, and AI builder. Moderating the conversation is Nick Thompson, CEO of The Atlantic.

In this episode, I sit down with my friend Rohit Krishnan - writer of the Substack newsletter Strange Loop Canon - for a hands-on conversation about what it actually looks like to build with AI agents today. Between us we're burning through tens of billions of tokens a month - I hit nearly 100 million in a single day this week - and we share what we're each running on our own machines. We dig into the quirks and surprising power of tools like OpenClaw, Claude Code, and Cowork, debate why AI remains stubbornly bad at good writing, and zoom out to ask what a world of trillions of agents might actually look like — and what economic infrastructure it will need.

In this episode, I'm joined by Jaime Sevilla, founder of Epoch AI; Hannah Petrovic from my team at Exponential View; and financial journalist Matt Robinson from AI Street. Together we investigate a fundamental question: do the economics of AI companies actually work? We analysed OpenAI's financials from public data to examine whether their revenues can sustain the staggering R&D costs of frontier models. The findings reveal a picture far more precarious than many assume; we also explore where the real infrastructure bottlenecks lie, why compute demand will dwarf energy constraints, and what the rise of long-running agentic workloads means for the entire industry.

A week before OpenClaw was released, I recorded a prescient conversation with Mustafa Suleyman, CEO of Microsoft AI and co-founder of DeepMind. We talked about what happens when AI starts to seem conscious – even if it isn’t. Today, you get to hear our conversation. Mustafa has been sounding the alarm about what he calls “seemingly conscious AI” and the risk of collective AI psychosis for a long time. We discussed this idea of the “fourth class of being” – neither human, tool, nor nature – that AI is becoming and all it brings with itself.