#AI Safety

The Berkeley computer science professor who wrote the most widely used AI textbook, who was among the first mainstream AI researchers to engage seriously with the alignment problem, and who has developed the most coherent technical and philosophical framework for building AI systems that are genuinely safe. The intellectual architect of the cooperative AI vision.

The full intellectual story of AI alignment research — from Norbert Wiener's early warnings to the technical work of the current era, from the philosophical foundations to the specific technical approaches. The most important unsolved problem in AI, and the people trying to solve it before it matters most.

The story of the March 2023 open letter calling for a six-month pause in AI development — who signed, who didn't, what it meant, and what it revealed about the state of AI governance. The brief, charged moment when it seemed possible that humanity might collectively decide to slow down before building more powerful AI — and what happened instead.

The full story of OpenAI — the founding mission, the commercial pivot, the release of ChatGPT, the board crisis of 2023, and what it means that the organisation most explicitly committed to humanity's benefit has become the world's most commercially successful AI company. The most consequential organisation in the history of artificial intelligence, and the complicated man who leads it.

The siblings who left OpenAI to found Anthropic, the organisation that has put AI safety at the centre of its commercial strategy. The Constitutional AI approach, the interpretability research, the billion-dollar fundraising, and the genuinely difficult question of whether safety and commercialism can coexist at the frontier of AI.

The co-founder of OpenAI, the architect of GPT, the chief scientist who voted to fire Sam Altman and then reversed course, and the founder of Safe Superintelligence Inc. The most technically gifted and most philosophically serious of the deep learning generation's leaders — and the figure whose trajectory tracks the field's own ambivalence about what it is building.

The Nobel laureate who spent fifty years building the technology he now warns against — the full story of Hinton's departure from Google, his public warnings about AI risk, and what it means that the person who did more than anyone to make the deep learning revolution possible now believes that revolution may have been a mistake. The most consequential change of mind in the history of AI.

The third Godfather of Deep Learning, who spent his career building the mathematical foundations of neural network learning and who has, in the years since the ChatGPT moment, become one of the most active and most credible advocates for AI safety research and AI regulation. Why the person who built so much of what AI is now believes that the field's progress requires governance it has been reluctant to accept.

He kept working on neural networks when everyone told him it was a dead end. He developed backpropagation, built the research group that produced AlexNet, and triggered the deep learning revolution. Then he quit Google to warn the world that he was afraid of what he had built. The full story of the most important AI researcher of his generation — and the journey from stubborn optimist to frightened sage.

He invented convolutional neural networks, built the first AI system to process real-world visual data at scale, and became the architect of Meta's AI empire. The most argumentative of the Godfathers of Deep Learning — and why his arguments, however combative the delivery, usually turn out to be right.

The third Godfather of Deep Learning — his foundational contributions to recurrent networks, attention mechanisms, and generative models, his evolution from pure researcher to the field's most prominent voice for AI safety and ethics, and why the man who spent his career building increasingly powerful AI systems has become one of the most persuasive voices for ensuring those systems are safe.