No.

Just because AIs run on computers doesn’t mean their thinking must share the qualities we associate with computers, any more than your thinking must share the qualities associated with biology and chemistry and neurotransmitters.

When humans didn’t understand biochemistry, they attributed the liveliness of life to an irreplicable “vital essence.” But reality is not made of mundane material sometimes animated by a magical life-force. Life is made of mundane parts.

We do not mean to degrade intelligence, however, when we say that it is made of mundane parts and that machines could do the same work. See our extended discussion about vitalism.

The heuristic “machines cannot compete with humans” was wrong when Kasparov predicted that a machine lacking human creativity could never beat him at chess; it was wrong when people thought that AIs could never draw pretty pictures; it was wrong when people thought that AIs could never chat conversationally. The human brain is an existence proof that physical matter really can implement higher forms of intelligence, sufficient for running a technological civilization; and the human brain is vanishingly unlikely to be the only way to do that work.

We’ll expand on this point in one of the online supplements to Chapter 3: Anthropomorphism and Mechanomorphism.

AIs are new, interesting, weird entities.

Airplanes fly, but they don’t flap their wings. Robot arms function without soft skin or red blood. Transistors work very differently from neurons, and DeepBlue played world-beating chess without the kinds of thoughts that went on inside of Garry Kasparov. This is the usual course of technology.

When we don’t understand flight or game-playing well, we sometimes imagine that the approach used by biology is the only possible approach that can work. Once we understand a field a little better, this turns out to be very wrong.

The work of steering a chessboard was done quite differently by DeepBlue than by Kasparov, and the work of steering the world at large will almost surely follow a similar pattern. As discussed in Chapter 2, AI looks like it’s already doing the work that it’s doing in a very different way than humans would — though this may be a bit harder to see when it uses its intelligence to imitate humans! In Chapter 4, we’ll explore how these differences are likely to lead to weird places, with serious consequences.