What Is Intelligence?" on Audible: Worth a Credit If You're Into AI
What Is Intelligence? by Blaise Agüera y Arcas is a really good book. I'm interested in AI and machine learning, and I liked how he starts with intelligence from a biological standpoint and works his way toward machines. The arc stuck with me: the beginning of life, the development of intelligent beings, and the jump from living organisms to machines.
Some links on this page are affiliate links. If you book through them I may earn a small commission at no extra cost to you.
I've read and listened to enough books on this subject that I'd hate to attribute an idea to the wrong author. I do remember a discussion about machines replicating the brain to derive intelligence, and I'm interested to see where that goes.
Agüera y Arcas is a Google VP and Fellow and the founder of its Paradigms of Intelligence research organization. He brings biology and computing together in this book, and there's a lot to take in.
Life, prediction, and machines
In "Origins," he looks at life in computing terms: living things store information, reproduce and change. He also spends time on symbiogenesis, which is when separate living things become dependent on each other and form something more complex. His example is the origin of complex cells containing mitochondria. He argues those mergers drive the big jumps in evolution.
Prediction connects that biology to intelligence. In "Survival," he uses bacteria responding to changing conditions to explore how a living system can use information about its environment to guide its behavior. Later, he extends that idea to brains and AI. His proposed definition is the ability to "model, predict, and influence one’s future". In the same passage, he argues that intelligences can form a larger symbiotic intelligence together.
That's a broad definition, and he takes it somewhere people will argue about. He makes the case that some modern AI systems have a claim to intelligence and consciousness, a view that goes against the grain.
Finding your way through the book
If a section goes by too fast on audio, you can jump to it in the free online edition:
- "Origins" and "Survival" cover life as computation, symbiogenesis, and prediction in living systems. An interlude follows on the prehistory of computation.
- "Cybernetics" and "Learning" move through feedback, early neural networks, and how learning relates to brains and machines.
- "Other Minds," "Many Worlds," and "Ourselves" turn to social intelligence, consciousness, and the brain. These sections get philosophical as well as technical.
- "Transformers" and "Generality" examine language models, attention, and the author's case for general machine intelligence.
- "Evolutionary Transition" closes with his view of AI as an evolutionary transition and considers its risks and economic implications.
If you want the part about replicating the brain, look at the "Homunculus" section of "Other Minds". It discusses a thought experiment in which biological neurons are gradually replaced with artificial ones that perform the same functions. The question is what happens to consciousness. It's a thought experiment, not a lab result.
Is it worth an Audible credit?
The audiobook runs 17 hours and 7 minutes, and he narrates it himself. That's a lot of listening, especially as he moves between different fields. If you enjoy AI and machine learning enough to pause and revisit an idea, a credit makes sense. If you want a short introduction to using AI tools, this is likely the wrong purchase.
The online text is free to read. Try a section there first, or keep it handy for the parts you want to revisit after listening.
I'm excited about where this technology is headed, how it can improve people's lives, and how it will affect them. If you're into AI and machine learning and don't mind taking your time with it, it's a good book.
If you want more
Life 3.0 by Max Tegmark explores possible futures shaped by AI, including questions about consciousness. The Coming Wave by Mustafa Suleyman with Michael Bhaskar focuses on powerful technologies and the challenge of keeping them under control.