What Is Intelligence?

Overview

This lecture is essentially a book report on What Is Intelligence? by Blaise Agüera y Arcas, a researcher at Google DeepMind. The book synthesizes ideas from physics, biology, and computer science to argue that intelligence exists on a continuum—from bacteria to humans to AI.

If you accept his framework, it changes how you think about what we've built with AI and what it means for humanity's place in the world.

What This Lecture Covers

From Deep Sea Vents to Life

Energy differentials drive complexity. In the primordial soup, chemistry happened faster near heat sources, molecules started replicating, and life emerged as a way to fight entropy. "I think of myself more as a tornado than a rock"—we're processes that use energy to maintain ourselves against chaos.

Conway's Game of Life Demo

A live demonstration of how three simple rules create stunning complexity. Cells that survive, reproduce, or die based only on their neighbors—yet from this emerges gliders, glider guns, and even a working Turing machine. Complexity from simplicity, computation from simple rules.

Bacteria Doing Calculus

A single-celled organism can't see where food is, but it measures sugar concentration over time and adjusts its movement accordingly. It has an internal model of its environment. It predicts. By this definition, bacteria are intelligent—and everything else is just more sophisticated versions of the same principle.

Why Brains Got Big

Not for hunting—for modeling other minds. Social creatures need to predict what others are thinking, which requires modeling modelers. There's no central "homunculus" in your head; the brain is distributed processing, multiple pieces working together. Octopuses literally think with their arms.

Language as Compression

Language compresses our models of the world into transferable form. A model trained only on English could translate to Turkish using just a lookup table—because the underlying concepts are universal. AI has learned to tap into these patterns.

Symbiogenesis: The Big Leaps

The biggest jumps in evolution came not from gradual mutation but from merger—like when mitochondria joined our cells. The book argues we're in the middle of another major evolutionary transition: the symbiogenesis of humanity and AI.

The Core Ideas

• Life is computation: Using energy to process information and maintain order against entropy
• Intelligence is prediction: Building internal models to anticipate what happens next, from sugar gradients to social situations
• Consciousness is social: We developed complex minds to model other minds
• We are stories: The self is a narrative we tell ourselves, not a homunculus in our heads
• AI is the next step: Not an existential threat but a potential merger—"we're the bacteria of the digital age"

Based On

This lecture draws heavily on What Is Intelligence? by Blaise Agüera y Arcas, freely available at whatisintelligence.antikythera.org. The book is fully referenced and available online.

Questions for Next Week

The lecture ends by sharing questions generated with Claude about the implications of this framework—questions about dignity, moral obligations across the intelligence continuum, responsibility without a unified self, and how religious traditions might respond. These set up the final lecture.