Frankenstein: Artificial Intelligence and the Problem of Evil


            Viktor Frankenstein had a vision, a dream, really. Could he create life? He left his hometown of Geneva, Switzerland for a local university. The world learned of his exploits in 1818. Mary Shelley tells his tale in Frankenstein.

            I reread the book every so often, and just completed another reading. I am gobsmacked all over again. It’s not the sorrow and loneliness that stun me. This reading I see the book anew. It can, and should, be read as a meditation on artificial intelligence and the problem of evil.

            First, a quick overview.

            Young Viktor is a young man of promise. He’s transfixed by the work of medieval alchemists, hoping to find in the material universe keys to unlocking life’s mysteries. When he arrives in Ingolstadt, presumably to study at the university founded there in 1472, his professors mock his naivete. The alchemists were frauds. But there is a way to understand nature, and, perhaps, to unlock its secrets. Viktor tuns to chemistry.

            He proves to be a brilliant student. He masters the science, but without losing the dream of understanding nature’s secrets. He turns to the creation of life, seeking to create life out of matter.

            He succeeds, but doesn’t tell his readers the secret – it’s too terrible to relate. He’s created a monster, large and grotesque in form, capable of sentience, speech and, we learn, the all-too-human instincts for love, understanding and companionship.

            Tragically, Viktor recoils in horror once his monster gains life. He cannot face what he has done. The monster wanders, overcome with sorrow by the cruelty of humankind. He approaches a family seeking love, and is met with scorn and fear. He reasons that only another being like himself will meet his need for companionship. He tracks Viktor down to demand that life again be created; make an Eve for my Adam. A good part of the story is about the consequences of Viktor’s reaction to the demand. (I won’t spoil the story be relaying it. But it’s good, eerily good, in fact.)

            The story is no mere Halloween thriller, although in the pre-pandemic days of trick-or-treating, hundreds of young Franksteins lurched from home to home in search of sweets. (I suspect not one in one thousand of these revelers knew that the monster had no name; his creator was Frankenstein. My introduction to the character as a child was steeped in this misunderstanding. I recall well constructing a model of the monster, thinking his name was Frankenstein.)

            Think on that for a moment: a sentient being without human connection, with no name, left alone to finds its way in the world. This being suddenly become like us, a mysterious transformation of mere matter into the stuff of humanity. What is it? What will it want? What will it do to get what it wants?

            That analog to the debate about artificial general intelligence is perfect. Rather that making life of mere matter, Viktor’s goal, AGI is shrouded in a different mystery: if the mind is a mere epiphenomenon of the brain, that consciousness, that defining characteristic of our species, is a mere matter of complexity. Solve the equation of the human psyche and you will become what Viktor Frankenstein was: the creator of something new, and, perhaps, terrifying.

            What is the sum and substance of consciousness? We don’t know. But we are capable of creating machines that perform some of the tasks that humans perform. This limited functionality is called narrow artificial intelligence.

            But we are more than the sum of the tasks we perform. We are animated by drives and desires, some conscious, others apparently not. What’s more, we can set over-arching goals, and we can learn new tasks. Indeed, one of the tests used for evaluating the intelligence of a machine is its ability to learn to do something for which it has not been programmed. There are breath-taking developments taking place as you read this in the area of machine learning.

            Will there come a time at which a machine effectively says to itself, either orally or internally – what’s the point of all this? Can a machine have purpose? Does it even make sense to speak of human intelligence that is disembodied? How much of our unconscious response to the world anchors and directs our minds? What’s more, how much of our experience of purpose, of contentment, depends on our ability to reach beyond ourselves? (I’ve often thought that until a computer generates something like Augustine’s Confessions, there will be no real proof of general intelligence.) What if computers, much like Frankenstein’s monster, come to crave companionship with creatures much like themselves?

            Frankenstein is a chilling reminder of all that can wrong in our quest to become as God. It is a secular version of Paradise Lost, a meditation on sin, or evil, at the very heart of what it is to be human.

            The nature of evil mystifies. Does it exist in and of itself, or is it the mere absence of good? Or perhaps it's misdirected love? We can’t say what it is for ourselves, yet we toy with creating immeasurably powerful things that may well struggle to answer these same questions. Beware the beasts we create, Frankenstein warns.

            Will our machines rebuke us, as the monster did Frankenstein: “Cursed, cursed creator! Why did I live?” Or, “Increase of knowledge only discovered to me more clearly what a wretched outcast I was. I cherished hope, it is true, but it vanished, when I beheld my person reflected in water, or my shadow in the moonshine,…” Or again, “Unfeeling, heartless creator! you (sic) endowed me with perceptions and passions, and then cast me abroad an object for the scorn and horror of mankind.”

            Vengeance is mine, saith the monster. Will the machine speak thus? Will it be capable of experiencing the transformative power of grace?

            “In my view, the greatest risk to humanity’s potential in the next hundred years comes from unaligned artificial intelligence,” writes Toby Ord, in The Precipice: Existential Risk and Future of Humanity. Ord is referring to an intelligence untethered to the ordinary tasks of daily living, to embodied intelligence in human form, in a word, to AGI.

            Frankenstein is a warning to be careful what we wish for. We may unlock secrets only to expose truths too terrifying to endure.

            And you though the pandemic was a challenge?

            Happy Halloween, all. Read Frankenstein. It is a dark, all-too-human, tale of desire, knowledge and worm gnawing at the core of the human psyche.

(If you find this site interesting and would like to support it, please consider supporting it at Patreon: <a href="https://www.patreon.com/bePatron?u=43627650" data-patreon-widget-type="become-patron-button">Become a Patron!</a><script async src="https://c6.patreon.com/becomePatronButton.bundle.js"></script>)

           

Comments: (2)

  • Frankenstein
    There’s a reason the subtitle is “The Modern Prometheus.”
    Same Greek myth only in the world of modernity.
    Posted on October 28, 2020 at 7:02 pm by Nat
  • Machines will never be smarter than the humans that create them.
    Period.
    Posted on January 8, 2021 at 8:41 pm by Frank

Add a Comment

Display with comment:
Won't show with comment:
Required:
Captcha:
What is the month?
*Comment must be approved and then will show on page.
© Norm Pattis is represented by Elite Lawyer Management, managing agents for Exceptional American Lawyers
Media & Speaker booking [hidden email]