Pasquinelli, Matteo. The Eye of the Master: A Social History of Artificial Intelligence. London New York: Verso, 2023.
There’s no shortage of books about artificial intelligence right now. Most fall into one of two camps: breathless optimism or existential dread. The Eye of the Master is something else entirely—a rare and rigorous exploration of how we got here.
This is one of the most illuminating books I’ve read on the current wave of AI. Pasquinelli doesn’t waste time on hype or speculation. Instead, he takes us back to the intellectual roots of machine learning—tracing how ideas from neuroscience, cybernetics, psychometrics, and even art history quietly shaped the architecture of neural networks and the metaphors we use to describe them.
His central argument is deceptively simple: modern neural networks are not so much models of intelligence as they are assemblages of borrowed metaphors. Half-inspired by the nerves behind the eye and half-shaped by statistical techniques from psychometrics, these systems are better understood as tools for classification and compression than as steps toward artificial general intelligence.
This, to me, is the book’s greatest contribution: it clarifies just how parochial today’s AI systems really are. Despite the lofty language we use—learning, reasoning, vision—these models are optimized for narrow tasks and trained on vast datasets to perform well in constrained environments. They work because they bypass the messy richness of human cognition, not because they replicate it.
In that light, Pasquinelli’s historical perspective has sharp contemporary relevance. It reveals how much of AI’s power is not technological but conceptual—a sleight of metaphor that lets us mistake pattern recognition for understanding, and optimization for thought. That doesn’t bode well for AGI, at least not via current approaches. Scaling up architectures inspired by partial models of the retina and 20th-century psychometric tests is unlikely to produce anything resembling a mind.
That said, a word of caution: while I enjoyed this book tremendously, it’s not an easy read. Pasquinelli moves fluently across domains—from Plato to Peirce, Turing to Turing tests—with a density of argument that borders on the academic. For those with no experience with critical theory or the history of science, it may feel like heavy going at times. But the reward is worth the effort: a deeper, more nuanced understanding of what AI is really doing, and why we should be careful about how we talk about it.
If you’re looking for a clear-eyed map of how AI came to be what it is—not what it might become—The Eye of the Master belongs on your shelf.