Quick Take
- Narration: George Newbern brings a measured intellectual authority to Bennett’s ambitious synthesis, handling the shift between evolutionary biology, neuroscience, and AI with consistent clarity across twelve hours.
- Themes: Brain evolution across 600 million years, the five breakthroughs of intelligence, gaps between current AI and human cognition
- Mood: Expansive and intellectually charged, with a sense of genuine discovery in the science
- Verdict: A synthesis that earns its Sapiens comparison in scope while doing real scientific work, the kind of book that reshapes how you think about both brains and artificial intelligence.
I finished A Brief History of Intelligence over a long rainy weekend, and by Sunday evening I was looking at my own habits of mind differently. That shift, the feeling of being changed in how you understand something you thought you already understood, is the benchmark I use for science writing that is doing real work rather than just explaining findings impressively. Max Bennett’s book crosses that threshold, and it does so on a genuinely demanding subject: the 600-million-year evolutionary history of the brain and what it tells us about the specific capabilities and specific failures of modern artificial intelligence.
The Sapiens and Behave comparisons in the book’s own marketing are not wrong, though they flatten what is most distinctive about Bennett’s contribution. Yuval Noah Harari traces the sweep of human history with narrative energy; Robert Sapolsky explains human behavior through the lens of biology with rigorous specificity. Bennett is doing something different from both: he is using the evolutionary history of brains as a diagnostic framework for understanding what current AI systems can and cannot do, and why. That dual orientation, looking backward through evolution and forward through AI development simultaneously, is what makes the book feel genuinely original rather than a synthesis that recombines existing popular science books.
The Five Breakthroughs as an Organizing Frame
Bennett identifies five pivotal evolutionary leaps in brain development, each one conferring capabilities that prior nervous systems lacked and that modern AI systems have either replicated, approximated, or continued to struggle with. The first breakthrough involves basic stimulus-response steering in primitive organisms. Later breakthroughs introduce more sophisticated functions: the capacity to model the world rather than just react to it, the ability to simulate future states, the development of emotion and motivation as computational mechanisms rather than mere feeling states, and finally the emergence of language and culture as a distinct layer of intelligence that operates above and alongside the biological brain.
What Bennett does with this structure is more than taxonomy. Each breakthrough becomes a lens for asking what current AI systems have actually managed to replicate and where the replication falls short. The famous asymmetry the synopsis invokes, that AI can defeat a chess grandmaster but cannot load a dishwasher effectively, is not treated here as a cute paradox. It is treated as a diagnostic clue about which evolutionary breakthroughs have been successfully modeled in machine learning architectures and which have not. That diagnostic is the book’s most original contribution, and the chapters that work through it are the strongest in the book.
Bennett’s Unusual Credibility Position
One of the interesting facts about this book is that Bennett is an AI entrepreneur, not a neuroscientist. He has built AI systems professionally while educating himself deeply in evolutionary neuroscience, and the book has been endorsed by prominent working neuroscientists. Reviewer Steven Miller from SMU Singapore noted that Bennett has a very unusual perspective by virtue of combining applied AI experience with serious knowledge of neuroscience, and that combination is exactly what makes the book work. A pure neuroscientist writing this book might produce a more technically precise account of brain evolution but a less interesting analysis of where AI development stands. A pure technologist would produce a more current account of machine learning architectures but a less rigorous evolutionary foundation. Bennett is operating in the gap between those positions, and the gap turns out to be where the most interesting questions are.
Reviewer R. Michaels, who rated the book four stars, described it as a well-researched explanation of the 600-million-year evolution of brains, organized around five breakthroughs, where each chapter revealed something new. That description captures the reading experience accurately. The book does not assume you know which evolutionary innovations were pivotal; it builds the case from first principles and then uses the case to analyze the present. For readers without a neuroscience background, the curve is manageable. For readers with one, the synthesis is where the value lies rather than in any individual fact.
What George Newbern Brings to Twelve Hours of Science
George Newbern has narrated everything from thriller fiction to self-help, and his range shows in how he handles A Brief History of Intelligence. The challenge with this kind of narrative science is maintaining intellectual energy across a long arc without letting the performance become declamatory. Newbern’s instinct is always to serve comprehension over drama, which is the right call here. The chapters that describe specific evolutionary organisms and their neural capabilities could easily tip into a BBC documentary affect, and Newbern keeps them grounded and conversational. The AI-focused chapters, which are more analytical than narrative, land with enough precision that the argument is trackable even in a car or on a walk. The supplemental PDF mentioned in the Audible listing is available alongside the audio and worth having for any diagrams or figures that accompany the text.
Where the Book Is Most Vulnerable
The comparisons the book draws between evolutionary neural mechanisms and specific AI architectures are illuminating but sometimes run ahead of the precision that specialists in either field would be fully comfortable with. Bennett is explicit that he is writing for a general audience and that his synthesis involves simplification, but readers with working knowledge of deep learning may occasionally find the AI analysis less technically precise than the neuroscience content, which is itself a synthesis rather than a technical treatise. This is a book for people who want to understand the landscape, not for people who want to evaluate the specifics of any particular machine learning architecture.
Who Should Listen / Who Should Skip
Essential for anyone curious about the relationship between biological and artificial intelligence, regardless of technical background. Particularly valuable for readers in AI, cognitive science, or neuroscience who want a narrative synthesis of evolutionary brain development rather than a specialist text. Less suited for readers looking for a technical assessment of current AI capabilities or for those who have already read extensively in both evolutionary neuroscience and machine learning literature.
Frequently Asked Questions
Do you need a background in neuroscience or AI to follow A Brief History of Intelligence?
No specific background is required. Bennett writes explicitly for a general audience and builds the evolutionary framework from accessible foundations. Readers with background in either field will find familiar material in the individual strands, but the synthesis and the cross-domain analysis are where the book’s value lies, and that is accessible regardless of prior knowledge.
Is the AI content in the book current enough to be worth reading, given how fast the field moves?
Bennett is not writing a survey of current AI capabilities, which means the book has more durability than a technology overview would. His argument concerns structural questions about what evolutionary mechanisms AI systems have and have not successfully replicated, and those structural questions remain relevant even as specific model capabilities evolve rapidly. The book was published with endorsements from working neuroscientists and reviewed positively by readers with AI backgrounds.
How does George Newbern handle the transition between evolutionary biology content and the AI analysis sections?
Newbern maintains a consistent intellectual register across both, which is the right approach for a book that is arguing for the unity of these two domains. The transitions are handled without dramatic gear-shifting, which keeps the listener tracking the argument as a continuous whole rather than experiencing the book as two separate topics patched together.
The book mentions a supplemental PDF. What does it contain and is it necessary for the audiobook experience?
The supplemental PDF accompanies the audiobook in your Audible library. Based on the nature of the content, it likely includes diagrams illustrating the evolutionary timeline and brain development comparisons referenced in the text. It enhances rather than replaces the audio, and the listening experience is complete without it, though visual learners who want to trace the five breakthroughs visually will benefit from downloading it.
