Water-Based AI?

A Critique of Denis Noble’s Biological Chauvinism

Recently, Oxford biologist Denis Noble claimed that true intelligence must be water-based because water uniquely enables fluidity, flexibility, and randomness—traits he deems necessary for genuine intelligence. While poetic, this assertion fundamentally misunderstands the nature of intelligence and computation.

Noble argues:

"Real intelligence isn’t just fast computation—it’s fluid, flexible and fuelled by randomness. That's why all living organisms are water-based. Water is a virtually unlimited source of the random motion which drives creativity, consciousness, and thought."

Let's unpack why this reasoning fails:

1. Confusing Substrate and Computation
Intelligence, at its core, is substrate-independent computation. Noble's assertion mistakes the properties of water molecules (flexibility, randomness) for the computational properties necessary for intelligent behavior. Computation is fundamentally about information processing, decision-making, and adaptive pattern-recognition, none of which inherently require a watery medium.

2. Randomness Isn't Exclusive to Water
While randomness indeed plays a vital role in creativity—because creativity is fundamentally an evolutionary process dependent on variation and selection—the particular molecular randomness of water isn't special. Computational randomness can equally arise from quantum fluctuations, electronic noise, radioactive decay, or even lava lamps. In fact, digital systems routinely utilize such sources of randomness for robust cryptographic and computational purposes.

3. Biological Chauvinism
Observing that terrestrial life is water-based and extrapolating this fact as universally necessary for intelligence is unjustified. It assumes biological life on Earth defines the boundary conditions for intelligence everywhere. This leap ignores silicon-based digital intelligence already demonstrating creativity and adaptive learning without any watery medium.

4. Practical Universality of Silicon-Based Computation
Silicon-based systems are practically universal computers, capable in principle of simulating any computable process given sufficient resources. While actual computers face constraints such as memory, speed, and energy, no compelling theoretical reason limits silicon (or any other non-biological medium) from performing the necessary computations underpinning intelligence.

Conclusion
Noble’s argument has aesthetic appeal but no scientific rigor. Creativity and intelligence depend on evolutionary processes driven by variation and selective retention—not on the particular physical characteristics of water. Genuine randomness and flexibility, the true computational requirements of intelligence, are accessible through numerous non-biological mechanisms.

We can appreciate Noble's poetry, but it shouldn't be mistaken for sound logic.