What does AI have to do with the Enlightenment?
AI is a wager that increasing the speed and scale of engineered intelligence can yield human progress. We'll need a renewal of the Enlightenment's legacy if it is to pay off.
I’m reproducing my latest Substack essay here as part of my Wednesdays series.
When I told friends I had been invited to a gathering of The Re:Enlightenment Project at Cambridge University—scholars who have spent more than a decade asking how the Enlightenment’s legacy speaks to the present—to discuss AI, they were puzzled. What does the Enlightenment have to do with AI? The Enlightenment is a chapter in a history textbook; AI is remaking how we think, work, and imagine the future. What could they possibly share? Everything, it turns out—once we ask what kind of intelligence we are building, and what we want to do with it. These are Enlightenment questions.
Time without a Future
The Enlightenment was a watershed moment in the history of intelligence: for the first time, humanity would treat time not as a fixed condition to be endured, but as an open-ended horizon of progress and improvement.
Writing in 1784, Kant defined the age of Enlightenment as an ongoing relationship to the future, not a finished state. We do not ‘at present live in an enlightened age’, he wrote, ‘but we do live in an age of enlightenment’. By deliberately cultivating our intelligence individually and collectively, we could indefinitely improve life on Earth. It was, and remains, one of humanity’s greatest wagers.
That wager is now in doubt, and the Enlightenment is in danger of becoming merely a historical episode. When we stop orienting toward a better future, we lose the very capacity to imagine one—we are navigating blindly.
The US Constitution, signed in 1787 and ratified the following year, is an Enlightenment document. It calls into being a people and its nation as having an open-ended, amendable, and optimistic future. ‘In Order to form a more perfect Union’ is thoroughly an Enlightenment constitution of time: perfection is never complete, the future is a field of possibilities, and we must actively constitute it as open.1 Yet we haven’t updated it in any substantive way since 1971, when we lowered the voting age to 18.2
We haven't achieved perfection, and the explanation runs deeper than polarization and factionalism: we've lost the will to constitute time as having a future, and the ossification of the Constitution is Exhibit A.3
Can we collectively re-engage the future?
The last substantive amendment to the US Constitution was in 1971 when we extended the right to vote to citizens 18 years of age or older.
AI and the Enlightenment
AI is an enormous wager on the future, but it arrives as the Enlightenment’s project of collective progress is eroding. This is the source of its danger, and ours. Without an historical compass that aligns the growth of intelligence with the betterment of the human condition, AI narratives will oscillate between apocalyptic fear and utopian abundance.
Without that compass, we will lose the wager that makes progress possible—that we can face big problems and devise intelligent solutions even as each solution spawns new problems.
Can we hold both wagers together—AI’s and the Enlightenment’s—and let each inform the other? Or will AI slip into Silicon Valley’s more familiar modes of time without a future: addictive algorithms, advertising-driven business models, and faster delivery of our dishwasher pods?
Is our time becoming unintelligent?
Time without a future is unintelligent, and our innovations will be disoriented and disorienting so long as we remain so. It treats the present as determined by forces beyond our control—what our long history has always called fate. To lose the Enlightenment perspective on history at a moment when AI enters the scene is to play a risky game.
Our critical philosophical tradition has been sounding this warning since at least 1882, when Nietzsche diagnosed time without a future as the death of God—the dissolution of a willingness to believe in a larger power that guides us through history. ‘What did we do when we unchained this earth from its sun? Where is it moving to now? Where are we moving to? Away from all suns?’4
When Jean-François Lyotard wrote The Postmodern Condition in the late 1970s, he named what Nietzsche could only anticipate: our growing incredulity toward ‘grand narratives’—the overarching stories that orient our pursuit of knowledge to a better collective future. Without bigger stories, our collective intelligence becomes untethered from any commitment to improve the human condition. Lose these stories, and time risks collapsing into a present that cannot transcend itself.
Writing in 2024, Marina Garcés calls our condition ‘time without a future’—an embrace of the present as unbreakable necessity, the blind fate of political, legal, and cultural systems that refuse to be updated. While each of us may hold our own optimism or pessimism about individual prospects, the collective present is stuck.
The Enlightenment refused this vision of time. We should too.
Time with a Future
Time Out of Joint has been tracing how we constitute time and intelligence through its literary artifacts—from Achilles contemplating his two fates, to Sisyphus condemned to a future without time, to Babylonian astronomers reading the heavens as preventable omens. Augustine’s Confessions is another such artifact—one that speaks directly to the stuckness of time without a future.
Why bring Augustine into an essay about the Enlightenment? For Kant, the Enlightenment begins when each of us musters the courage to overcome our own intellectual limits. Confessions (ca. 400 CE) may be the first sustained biographical narrative in the Western tradition to do precisely that.
His famous story of stealing pears captures time without a future with lasting accuracy. He describes ‘having no cause for wrongdoing save my wrongness’:
The malice of the act was base and I loved it—that is to say I loved my own undoing. I loved the evil in me—not the thing for which I did the evil, simply the evil: my soul was depraved, and hurled it down from security in You into utter destruction, seeking no profit from wickedness than to be wicked.5
Augustine narrates an urge consumed by itself. It doesn't even want the pear as a future pleasure. It wants only its own indulgence—an abyss that has no exit and no tomorrow. Time cannot transcend the now of this urge. The future is swallowed whole. Augustine is describing non-time.
A grander narrative, however, re-orients the urge toward the future. Writing more than a thousand years before the Enlightenment, Augustine already understood its wager: our actions gain direction and meaning within a larger story—in his case the Christian narrative of creation, fall, and redemption. The grand narrative provides what non-time cannot: a critical distance from the present, and an orientation toward a better future.
Confessions models a self that can create a future that is not simply more of the same: ‘This then is the fruit of my confession … not what I was, but what I now am, and what I continue to be’ (10.4.6).
It is a key moment in the history of intelligence.
Toward a History of Intelligence
Time must break open from this self-consumption if we are to have a better future. But the concomitant danger is the single-threaded narrative—oppressive at scale, and violent when those in power try to enforce conformity. That too is a lesson of the Enlightenment.
My contribution to a new Enlightenment is to treat intelligence not as a thing but as an expanding, adaptive capacity to transform necessity into possibility. Intelligence, therefore, has a history legible through the artifacts. From cave paintings to astronomy, from Confessions to LLMs, there are many stories. Telling them may be among the most urgent undertakings if we are to restore a future to time and continue to be, as Kant desired, an age of Enlightenment.
The Magna Carta (1215) was an inspiration for many of the US founders, but it is a contractual document between existing nobles and the existing monarch who make up an existing feudal kingdom, which we anachronistically could call a nation. It is not the founding of a nation with an open-ended future and an indefinitely growing citizenry.
The 27th Amendment of 1992 restricted Congress from immediately changing its own compensation. It hardly counts as a substantive expansion of rights or an opening of any new horizon of possibility.
‘We do not think about bettering ourselves but only about obtaining more and more privileges in a time that is going nowhere because it has given up on aiming at a better future.’ Marina Garcés, New Radical Enlightenment, Verso: 2024, page vii.
The Joyful Science §125, ‘The Madman’. Adrian Del Caro translation.
Confessions, 2.5.9, F.J. Sheed translation.