The Faust Myth Foretells AI
The technology offers a devil’s bargain that we should accept with care.
Given that AI is here to stay, it behooves those of us attempting to adapt universities to the realities of the mid-21st century to pay attention to insights gleaned from past wisdom as we develop our curricula. Will we turn our backs on AI, pretending that device-free exam rooms will somehow prepare students for an AI-saturated world? Will we simply allow our students to sell their figurative souls to the AI devil, allowing it to do much of their thinking for them? Or will we recognize that today’s most important skill is to think alongside machines and start building institutions bold enough to teach students how to do this?
The most up-to-date AI models are more knowledgeable not only than any individual human mind but are in certain respects able to rival entire expert communities. They provide an interactive, pattern-based generator enabling them to do enormous labor. Given their novelty, one might suppose that ethical issues surrounding their adoption would be a purely contemporary concern. Nevertheless, works of literature frequently debate significant issues long before humans create the technology that allows them to be instantiated. So it is with AI, one of whose leading purveyors in world literature turns out to be not Sam Altman or Elon Musk but a rather more complex figure—Satan.
The earliest major literary work to develop the idea of a deal with the devil in the context of AI is Christopher Marlowe’s The Tragical History of Doctor Faustus (early 1590s). Today’s reader is struck by Faustus’ situation before he sells his soul. He is no desperate second-rater, but a respected scholar. Describing his accomplishments, Faustus points to his skill as a physician: “Are not thy bills hung up as monuments / Whereby whole cities have escaped the plague / And thousand desperate maladies been eased?” But what can be achieved through purely human effort is insufficient. He turns to magic and signs over his soul to the devil in exchange for access to artificial intelligence for twenty-four years. Unfortunately, Faustus does not employ his gift to make the world better but instead fritters it away in frivolous adventures and dies miserably.
Marlowe pinpointed a central issue surrounding the adoption of AI. The AI user can know everything, but there is a danger that in accepting this tool he may give up his duty to do the work (intellectual/moral) that would give meaning to this knowledge. Merely ask Mephistopheles or ChatGPT and nothing further is needed. In the end, however, sloth kills, and the devil will have his due. Marlowe asks: If humans no longer need to work to achieve knowledge, will they still value it or will they spend their lives playing the equivalent of elaborate video games?
Goethe’s version of Faust—manifesting in literature two hundred years after Marlowe’s—is well respected by his scholarly colleagues, but after a life of study he despairs, “All I see is that we cannot know!” He craves endless knowledge without work:
Therefore I have turned to magic,
so that by the spirit’s might and main
I might yet learn some secret lore;
that I need no longer sweat and toil
and dress my ignorance in empty words;
that I might behold the warp and woof
of the world’s inmost fabric,
of its essential strength and fount
and no longer dig about in words.
Mephistopheles promises precisely the ability to gain insight without toil. “My friend, in this one hour you will gain / far more for all your senses / than in a year’s indifferent course.”
Goethe’s hero is willing to sign over his soul but fails to make good use of his newfound powers. Instead, like many of our students, he decides to apply his newfound knowledge to a field in which he has no expertise and one for which the tool is singularly unfit—in this case, love. The result is disastrous, for Faust himself and for the object of his seduction and her family.
However, the Faust story is quite elastic, and 20th-century authors, including Thomas Mann in his 1947 novel Doktor Faustus, would revisit it to illustrate complexities that could provide insights for today. Mann’s hero, Adrian Leverkühn, is a talented composer who dreams of achieving transcendental greatness. The novel’s key chapter records a conversation between the composer and the devil, who materializes suddenly and offers Leverkühn twenty-four years of time. But, unlike earlier devils who purvey knowledge without the need for work, and in a twist on the usual facile Faust-narrative of soul-selling, Mann’s incarnation promotes AI as a tool to be deployed alongside human ingenuity to allow for the production of compositions that would otherwise be unrealizable. As the devil puts it in his baroque idiom:
Time? Simple time? No my dear fere, that is not the devyll’s ware. For that we should not earn the reward, namely that the end belongs to us. What manner of time, that is the heart of the matter! Great time, mad time, quite bedivelled time, in which the fun waxes fast and furious with heaven-high leaping and springing.
Leverkühn accepts the devil’s bargain and, if we are to believe the novel’s narrator, during the twenty-four years at his disposal he creates a series of brilliant compositions that set the agenda for the development of music for generations. He is able to accomplish this, however, because years of training had given him enormous expertise and because he remained willing to spend endless hours refining his gifts with the help of Satan’s AI boost. True to form, at the end of the allotted period Leverkühn burns out and dies, but we can suppose he found his deal with the devil to have been fair and fulfilling.
As educators develop our university curricula in the AI age, our wisest choice out of the trinity offered to us by classical literature is to attempt to follow Mann’s lead, recognizing that in the mid-21st century the challenge will be not to choose between human or artificial intelligence but to discover a synthesis through systematic experimentation—one that might allow our students to produce transcendental work like those of Leverkühn, though hopefully without the need to burn out in the end.
Andrew Wachtel is president of inVision U, a university project piloting in Kazakhstan and featuring a curriculum allowing students to synthesize human and artificial intelligence.
Follow Persuasion on Twitter, LinkedIn, and YouTube to keep up with our latest articles, podcasts, and events, as well as updates from excellent writers across our network.
And, to receive pieces like this in your inbox and support our work, subscribe below:
The agricultural revolution slowly changed the lives of countless would-be hunter gathers in ways that were not and could not be predicted. Fast forward 10K years and we have the industrial revolution that reduced the need for farm and domestic workers while at the same time providing jobs in industry. Work in the factories early in the industrial revolution was hard and at times brutal, but wages put actual money in people's pocket, which in of itself was liberating. Now we arrive at the doorstep of what will be an AI revolution. Those who think they can throw their sabots into the gears of the AI revolution to slow things down are misinformed. Those who think they can predict the outcome of the AI revolution are delusional.