Share E-Book
Scan to open this page

Scan with your phone to open this page

Author: Tomaso Poggio, Marco Magrini

Rating 3.5 AI initial rating

An exploration of the past, present, and possible futures of AI from an AI pioneer. Intelligence remains an open mystery. The fact that only humans possess language, alphabets, and science doesn’t grant us a monopoly on it. We share our existence with millions of other species, both animal and plant, each endowed with such a diverse range of cognitive abilities that they form an almost infinite spectrum of intelligences. Brains Minds Machines explores these intelligences and goes further to examine the possibilities, risks, and opportunities of our newest one—artificial intelligence. Thanks to the combined emergence of more sophisticated algorithms, vast databases, and immense computing power, the age-old aspiration of mathematically replicating human intelligence has achieved unexpected milestones. While still far from truly succeeding, a small “zoo” of artificial intelligences can already perform numerous tasks typically done by humans. Like electricity or computers, this technology is poised to transform society, the economy, and daily life, bringing with it a host of risks and opportunities. However, it is also going to be an opportunity for new, extraordinary scientific discoveries, starting with the very secrets of intelligence itself —the fundamental principles common to all forms of intelligence.

AI Reading Assistant

Whole-book reading guide from stratified index samples; jump to passages in the text

AI guide
【One-Line Pitch】 A first-person tour of intelligence in all its forms—from bacteria and birds to the human brain and the machines we are building to imitate it—written by an AI pioneer who wants readers to think in questions rather than certainties. Best for curious general readers and technically minded newcomers who want historical and biological context before the hype. 【Book Arc】 - **Opening (~0%–10%)**: Frames the whole project—intelligence is not a human monopoly—and sets the book's method: a mathematician's first-person narrative, cautious about prediction, more interested in questions than theories. - **Early (~10%–35%)**: Traces the deep history of intelligence, from the physics of the early universe through simple cells, multicellular life, the dinosaur extinction, mammalian radiation, and the emergence of *Homo sapiens*, using a 1983 walk with Francis Crick as a narrative anchor. - **Middle (~35%–55%)**: Dismantles human-centered measures of intelligence—mirror tests, cortex size—and surveys cognition across animals (corvids, primates, cetaceans, insects) and the case for plant intelligence, introducing the *g* factor and its discovery in nonhuman species. - **Late (~55%–85%)**: Turns to artificial intelligence: how algorithms, data, and computing power produced a "zoo" of narrow systems, the economics and data politics around them, and the risks the authors consider more human than robotic. - **Ending (~85%–100%)**: Asks what law, theory, and science can offer—arguing that AI research may itself unlock the fundamental principles shared by all intelligences, closing with an epilogue on uncertainty. 【Key Takeaways】 - **Intelligence is a spectrum, not a ladder** (Early–Middle): The book's organizing claim is that millions of species possess diverse cognitive abilities, so ranking them by human yardsticks distorts what we are measuring. - **Evolution is a story of cooperation and error-correction** (Early): From quarks to cells to nervous systems, the narrative stresses that life advanced through molecular partnership and repair mechanisms—context that makes intelligence seem less like a miracle and more like an outcome. - **Standard intelligence tests are species-biased** (Middle): Mirror self-recognition and cortex size fail as universal metrics; dogs may not recognize themselves because they navigate by smell, and birds solve problems with a pallium rather than a cortex. - **The *g* factor is not uniquely human** (Middle): Spearman's general intelligence factor, once thought to separate humans from animals, has been found in rats, primates, corvids, and some dogs—undercutting human exceptionalism. - **Plants and simple organisms also process information** (Middle): The book extends the intelligence debate beyond animals, noting that all organisms perceive and respond to their surroundings, which challenges Aristotle's hierarchy of souls. - **AI's rise is a convergence, not a single breakthrough** (Late): Better algorithms, vast databases, and immense computing power together produced systems that perform many human tasks—yet remain far from genuine general intelligence. - **The greater risks are human, not robotic** (Late): The authors frame dangers around misinformation, power, and social disruption rather than a robot uprising—an argument worth weighing against more alarmist AI writing. - **AI may be the key to understanding intelligence itself** (Ending): The deepest payoff of building intelligent machines could be scientific discovery about the principles common to brains, minds, and machines. 【Reading Tips】 - **Read the early chapters as context, not prerequisites**: The evolutionary history is engaging but long; if you already know the biology, skim toward the animal-cognition and AI sections. - **Slow down on the *g* factor and animal-cognition chapters**: These carry the book's central argument and are where the authors most directly challenge common assumptions. - **Treat predictions skeptically—as the authors do**: The prologue explicitly warns that forecasting is unreliable; read the AI-economy and risk chapters as framing questions, not forecasts. - **Keep the Crick walk in mind as a through-line**: The 1983 conversation recurs as a narrative device connecting molecular biology, evolution, and intelligence—use it to orient yourself when the timeline jumps. - **Finish with the final chapter and epilogue**: They state the book's real thesis—that AI research is a route to understanding natural intelligence—more directly than the middle chapters. 【Coverage Limits】 The excerpts cover the prologue, the evolutionary and animal-intelligence chapters, and the table of contents for the later AI sections; detailed content from the economics, risk, and final science chapters is only partially represented, so this guide's late-book claims rest on chapter titles and brief passages.
Excerpt 1
77 Massachusetts Avenue Cambridge, MA 02139 -->mitpress.mit.edu --> © 2026 Massachusetts Institute of Technology Originally published as Cervelli Menti Algor...
View in text
Excerpt 2
them A, G, T & C. We now find that the pairs we can make . . . are only A with T and C with G.” “Now we believe,” he continued, “that the DNA is a code. That...
View in text
Excerpt 3
, us, who have multiplied to become more than eight billion. An Almost Infinite Gradation How do we know all that? Because countless brains—especially in the...
View in text
Excerpt 4
y 20 years ago, it was discovered that rats have a g factor. A little like in Spearman’s experiments, the rodents were given various tests, from the time it...
View in text
Excerpt 5
became essential for verifying our theories experimentally. Since Aristotle’s time, theories about the mind had been exclusively the realm of philosophy. Wit...
View in text
Excerpt 6
one computer language—have completely altered the landscape. Most probably, ChatGPT, Gemini, LLaMA, Claude, Perplexity, DeepSeek, and dozens of other generat...
View in text
Excerpt 7
ferent intensities, virtual synapses have different weights. Also, just like every biological neuron, the perceptron’s central neuron had its own characteris...
View in text
Excerpt 8
by Bell Labs, became their competitors. But that’s not all. The forecast made in 1965 by Gordon Moore, cofounder of Intel—now known as Moore’s law —that is,...
View in text
Tags
AI categories
Artificial IntelligenceScienceTechnology
ISBN: 0262054361
Publisher: MIT Press
Publish Year: 2026
Language: English
Pages: 224
File Format: EPUB
File Size: 409.5 KB