Skip to content
Coordinates of Thought
Sign in
Discussion

Does a deterministic program “choose”?

A chess engine searches millions of positions, scores them and plays the best move. We naturally say it "chose" the knight move. Yet it's fully deterministic: same position, same settings, same move every time.

Spinoza thought our sense of free will comes from being aware of our actions but ignorant of their causes. In one letter he imagines a stone in flight that becomes conscious and believes it moves freely.

Dennett, in Elbow Room (1984), argues the reverse: the kind of freedom worth wanting is the ability to evaluate options and act on reasons, and determinism doesn't take that away.

So is "choose" just a figure of speech for the engine? And if it's a figure of speech there, why isn't it one for us?

4 replies

  1. Emeka Nwosu

    Contributor

    From the physics side: at the scale of neurons, the brain is probably close enough to deterministic that quantum effects don't help. So either choice is compatible with determinism or nobody chooses anything.

    Helpful · 1
  2. Aoife Brennan

    Contributor

    I'd say the engine doesn't choose because it doesn't care about winning. Choice involves wanting something and being responsible for the result. The engine has a scoring function, not desires.

    Helpful · 2
    1. Pieter de Jong

      Contributor

      Aoife, I'm not sure "caring" is the line. Engines have something that functions like preferences. Maybe the real difference is that nobody holds the engine responsible, and that's a social practice rather than a fact about its insides.

  3. Henrike Vogel

    Contributor

    Kant's answer would be that we must regard ourselves under two standpoints. As objects in nature we are determined; as agents deliberating we can't help acting under the idea of freedom. The engine doesn't deliberate under any idea at all.

    Helpful · 3

Members reply here. Reading is always free.

Sign in to reply