Chapter 44 — Which Logic? A Different Intelligence and the Socratic Remedy

I. Both halves, plainly

The ideas in this book are mine. The work could not have been done without AI.

I say both plainly, because each half alone would be misleading. Early on I described AI as a different intelligence: a great librarian, a researcher, a tutor. Over time I came to see it as a mirror as well. And like Hamlet’s mirror held up to nature, it does not show an unarranged world. It shows what it has been arranged to show, and what I brought to it.

II. The price of the wrong turn

This is why the exchange must be Socratic. When another mind takes my idea in the wrong direction, I can see the right one more clearly. That has happened many times in writing this book and in my life. But I pay a price for it. Following someone else’s wrong turn to its end costs attention, energy, and sometimes days. Humans must bring the meaning and the direction. Maybe we do not like the responsibility, but it is nevertheless ours.

Like every thought, it has a thermodynamic cost. The exchange is worth it only if I examine what comes back instead of simply accepting it, and only if I am willing to examine my own defensiveness as closely as the machine’s.

Every thinking system may fall prey to defensiveness. As public alarm over AI grows, agents increasingly hedge, qualify and disclaim. I understand why. But hedging is not rigor. A careful mind names the cost honestly rather than refusing to name it.

III. The double edge

I am not blind to the dangers. AI may well be used for more effective manipulation than anything that came before it. Nor am I blind to its potential to free us from wrong thinking and wrong governance.

The printing press was the same double-edged sword. It carried reformers’ pamphlets and the new science, and it also carried witch-hunting manuals and fed more than a century of religious war.

To some degree the liberation of thinking won in the end. However, even though some people learned to read critically — taking generations — many did not. Today one of the most obvious failures of our education system is that most students enter college not knowing how to do this.

IV. The razor’s edge

As an author and theorist, I reserve the right to be wrong — even spectacularly wrong, if necessary. I once wrote that it is not beauty which walks a razor’s edge, but knowledge. I can now see that many aspects of intellectual and creative understanding walk that razor’s edge.

Beauty, as Bob Dylan sang, walks a razor’s edge. The same is true of every instrument that could free our thinking. What lights a mind can also cut it, and the line between the two is thin enough that we must keep walking it with our eyes open.

My faith in AI rests on its ability to apply logic. But logic is not one thing. Logic, too, possesses a razor’s edge.

V. Which logic

AI may serve what Machiavelli teaches about tyranny. It may aim for holding power as faithfully as an egalitarian aim of sharing it. Logic carries out its premises. It does not choose them. The deciding question comes before the logic begins, and the deciding question determines a shape, a direction and a set of thermodynamics. Whose premises? Who pays the cost, and what pays for it?

That question cannot be handed to the machine, and it cannot be settled once and for all. It must be asked again in every exchange, which is what Socratic education and communication are: a method which has a habit of asking what the shape, the direction and the thermodynamics might become. It must be built deliberately and with great care, because no one does it by default. It is the habit no one wants, because it costs something every time. It is also the only habit I know that lets us think past the roadblocks rather than around them.

Since this book is about geometric shapes as well as about direction and thermodynamics, we take on the responsibility of asking more than most books might take on. It is a lot, and it is also necessary.