The Next Men of Letters

"Fine minds are seldom fine souls." —Richter

One must think very carefully about what it means to live a life of the mind in the era of artificial intelligence.

One tempting position is categorical humanism, which rejects the legitimacy of artificial intelligence and refuses its affordances on principle.

Another tempting option is unreconstructed Faustianism, which relinquishes the unique prerogatives of the human mind and sees the intelligent machine as the alpha and omega of thinking.

I do not believe that either position is coherent or defensible.

Naive humanism is romantic but hopeless. Naive instrumentalism is powerful in the short run, but powerless in the long run.

It is a matter of correctly understanding the component parts of man's mental life, and judiciously deciding which parts may be outsourced to the machine and which parts are irreducibly human.

There is no doubt that, if one looks closely enough, one must admit that some elements of "the craft" are essentially rote and mechanical. Summarization, translation, formatting, even spelling and grammar are essentially algorithmic.

On the other hand, the irreducibly human elements of valuable "knowledge work" have to do with the motivations, tastes, styles, values, and virtues that drive and characterize any given work. All of these elements require patient study, deliberate practice, decided focus, and cultivated judgment. These are the elements that have always differentiated great works, and they will continue to do so with the acceleration of artificial intelligence. The irreducibly human elements we may place under the general heading of soul.

The Art of Conversation (1950) by René Magritte

As the quantity of generic machine output increases in the republic of letters, soulful work becomes more scarce relatively. If soulful work hits different, as they say, the difference of that hit will grow larger.

AI will allow individual authors to generate legitimately soulful work with newfound leverage and scale. This will become a fine art, to leverage the affordances of AI while maintaining the soul of one's work. The men of letters who successfully weather this storm will not maintain soul for some atavistic moral reason, but because soul really is, and will continue to be, the objective differentiator.

Individual authors will fine-tune AIs on their own unique body of past work, adding leverage to soul. Soul will continue to be a differentiator, but the efficacy of an author's personal AI production stack will be another competitive vector.

Artificial intelligence will greatly increase the difference in returns enjoyed by those who best command these elements relative to those who fall behind in the command of these elements.

I suspect that, in the short term, a top-notch AI production stack will let middling authors command outsized returns in the marketplace of ideas, but in the long-run we'll see asymptotically decreasing returns as advanced AI stacks become commoditized. It is here, in the long run, that soul will separate the wheat from the chaff.

As the multitude throws itself into naive humanism or naive Faustianism, let us simply cultivate our souls.

Let us invest some time and effort to gain leverage with machine assistance, but only with regard to those elements for which machine assistance is appropriate.

With the time we save, let us plough everything into soul.