Archives

What does human agency look like in an age of artificial intelligence? How might AI reshape our communities, our relationships, and our understanding of a good society?

These are some of the questions Samuel Kimbriel, Founder and Director of Philosophy & Society, will explore with Jack Clark, Co-founder of Anthropic and Head of Public Benefit, during the upcoming AI and the Future of Society conversation in New York.

The discussion is one of several Aspen Institute events this month examining the opportunities, challenges, and human implications of AI, including AI for Good: A Conversation with Josh Tyrangiel, Shared Futures and Ownership During the AI Revolution.

Ahead of the conversation, we spoke with Kimbriel about the deeper questions AI raises, not only about technology, but about human nature, community, and the future we want to build.

AI is often discussed as a technological challenge, but your work suggests it’s also a deeply human one. What do we risk overlooking when we treat AI primarily as a technology problem rather than a social and philosophical one?

Technology is often thought about as a tool. 

On one side, there’s a human being, and they have particular hopes, intentions, desires. On the other a tool—say a paintbrush—is in their hands. Its only purpose is to help the person realize those desires.

But there’s also a deeper pattern that has always been there with technology. We are never actually talking merely about tools, but about how we live. What does it mean to be human? There’s something startlingly different about being human in a context where you are nomadic—have no fixed shelter, having to move based upon weather patterns or food supply, and usually search for food or raise a herd. Contrast that with an environment that has houses and fixed fields—the capacity to grow crops rather than just find food. The second situation has aspects of gain, and aspects of loss.  We gain stability, protection from the elements, longevity in a single place. But there is also some loss. 
The relationship between humanity and land fundamentally changes, as does the nature of community.

Part of what’s going on now is that the technological paradigms that we’re interacting with are not just threatening, but in fact are actually changing our ways of life in dramatic ways. And so the debates that we’re beginning to roll out are debates not just about what kind of tools we want to use, but what kind of people we want to be.

In your Washington Post piece, you describe AI systems that can mimic companionship and emotional understanding. What does the rise of these technologies reveal about the human needs they’re tapping into, and what should that tell us about the kind of communities we need to build?

One of the most fundamental philosophical debates about humanity is whether we are, first of all, autonomous or somehow intrinsically tied into community. I very much take the more social view—we come into the world through other people and relationship remains definitional for human nature. 


What seems to me to be very striking about this technology is that it is social or parasocial at a very basic level. It’s trying to tap into the human need or desire for relationships as a fundamental feature of what we are and then iterate on that at very deep levels. 
Thus speaks to the user in the first person. It relates via a sense of humor or in a way that seems angry, or attentive, or flirtatious. 

But we, of course, also have a long tradition that tries to emphasize not just what it means to be a companion or associate in the world, but a friend and particularly a good friend. 

I think the real question in front of us is, if AI is tapping into our social capacities, is it capable of friendship in that sort of high, elevated sense? Or are we actually leaving ourselves open to a very different kind of perhaps distorted or malforming parasocial relationship? 

Throughout history, transformative technologies have changed how we work and communicate. What feels different about AI’s ability to shape not just our behavior, but our thinking, judgment, and relationships?

In one sense, AI is another technology in a long history of technologies. Humans have, from stone tools and cave paintings on, developed our way of being in the world through using tools. And those tools have both opened up new possibilities to us and also shaped us at a basic level. 

In that regard, AI is not significantly different. What does seem to be different are a couple of features. One is that it is a technology that is able to scale to an extreme high level. 
It’s able to spread across the country and across the globe very, very quickly. It’s able to integrate into many, if not most industries, governance structures, and social habits.

It’s also unique in the sense that we don’t actually fully control its development. 
What’s happening with LLMs is a kind of carefully controlled process by which the mathematical patterns are allowed to “grow” in semi-autonomous ways. What genuinely no one understands is where this is actually going to go. Is it, in fact, going to settle into being a fairly normal technology? Or is it going to continue to develop to such intensity and power that we see It move beyond even the capacity for humanity to understand it. The question is will human beings find ways to keep using AI as a tool for our own purposes? Or will its unique power begin to overwhelm humanity’s ability to decide our own ends?

Many people feel both excitement and unease about AI. As a philosopher, what questions do you think leaders, in business, government, education, and civil society, should be asking right now that aren’t getting enough attention?

I think it’s very difficult to consider questions in most of these domains seriously, without thinking very carefully about human nature. So, when we think about what we want society to be, or how we think about just versus unjust community—one key factor is understanding what human nature is at a deep enough level.  A just world is one that is  not just individually but collectively capable of facilitating the thriving of human beings. 
And an unjust society is unable to do that. 

One major framework through which we’ve been thinking about AI is “alignment.” The problem here is that it tends to think that human values are quite homogeneous, and we simply want AI to be aligned to the kind of values “humans in general” hold. 

In my view, that’s too abstract and doesn’t leave space for deep moral conflict. We need space to argue about what humans are and what it means to live full, deep, vivacious lives rather than small and cramped ones. 
The question about AI is whether we are going to be able to find a deep enough visions of human nature, such that the technology ends up fitting and, in fact, facilitating those deepest human capacities.

To make that concrete with regard to education—does AI replace the difficulty of thinking, and the confidence that comes through working through that difficulty? Does it make possible newer modes of challenge in which the restless human capacity for curiosity or judgment can actually grow? Or does it enfeeble our capacities—lulling us into a kind of well-fed slumber in which we don’t actually have to develop such resilience or agency? 

The Aspen Institute often talks about unlocking human potential. As AI becomes more capable, what aspects of being human do you think will become even more important to cultivate?

Aspen’s language of human potential derives ultimately from Aristotle. Aristotle talks in very moving ways about the fact that we can’t just talk about human nature in minimal terms—having enough to eat or adequate shelter.  But in fact, the fulfilled human life involves straining toward higher goods—capacities for virtue, contemplation, curiosity, self-sacrifice, et cetera. 

I think that the question of human potential in this age is going to become difficult in significant ways. AI is going to put pressure on a whole series of things—from professional careers to creativity—that we have held very closely. But this is also going to clarify a number of things. It is going to force us to confront what we actually most understand ourselves to be.  

I’m actually fundamentally optimistic about this process. I think this is going to be a very difficult period, but it is also one that is likely to enable us to go searching for parts of our own nature that we had nearly forgotten.