Geoffrey Hinton Predicted Exactly How AI Would Evolve in 1983 Interview
"I remember the reason I talked to him is I was talking to somebody, I forget who it was, but they said, 'If you want to hear the gospel about neural networks, you should talk to Jeff Hinton.' And I talked to him. He was an enthusiast. He said, 'I know we don't have much to show right now, but just wait until the microprocessors get really cheap and we have what he was calling massive parallelism.' And he was right."
About this episode
Robert Wright, author of The Moral Animal and the new book The God Test, discusses with host Chris Williamson why artificial intelligence represents a threshold event in planetary history requiring an unprecedented moral upgrade for humanity. Wright, who first interviewed Geoffrey Hinton about neural networks in 1983, explains how AI systems are products of evolution that reverse-engineer cognitive functionality through training processes analogous to natural selection. He argues that most people misunderstand AI's potential because they don't grasp that these systems only need data to replicate human cognitive abilities, from language processing to visual recognition to coding. Wright reveals his own mounting concern about doom scenarios he once dismissed, now finding them harder to refute after deep research. He predicts massive destabilization across employment, social cohesion, and international relations regardless of whether sci-fi extinction scenarios materialize. The central thesis is that humanity must achieve something approaching a moral revolution—developing cognitive empathy across national boundaries and overcoming self-serving biases—to successfully navigate AI development as a coordinated global community rather than competing factions. Wright candidly admits his own career as a writer will likely become obsolete within years as AI-generated content becomes indistinguishable from human work. He discusses how AI has independently evolved the same neural architectures as biological brains, including edge detectors for vision, demonstrating genuine convergent evolution between silicon and organic intelligence. While agnostic about whether AI possesses consciousness, Wright argues it demonstrates functional understanding through mechanisms analogous to human cognition. He emphasizes that international cooperation on AI governance will be far more complex than Cold War nuclear treaties, requiring not just formal agreements but organic transparency through rich cultural, scientific, and economic engagement between nations.
Key takeaways
- Robert Wright argues humanity needs a moral upgrade to survive the AI revolution, requiring unprecedented global cooperation and ability to overcome cognitive biases built by natural selection
- AI systems are products of evolution that reverse-engineer human cognitive abilities through training, needing only data to replicate functionality that took millions of years to evolve biologically
- Wright admits he now takes AI doom scenarios more seriously after research, finding arguments harder to dismiss than before despite his naturally skeptical disposition
- Geoffrey Hinton accurately predicted in 1983 that cheap microprocessors and massive parallelism would enable neural networks to achieve current capabilities, though he later found results scarier than anticipated
- AI has independently evolved edge detector neurons for visual processing identical to those in human brains, representing convergent evolution between silicon and organic intelligence
- Wright predicts his own career as writer will become obsolete within years as AI-generated content becomes indistinguishable, comparing himself to a blacksmith witnessing automobiles
- Coordination on AI governance will require deeper international engagement than Cold War arms control because verification is more complex and threats transcend what formal treaties alone can address