Civic education in an age of AI

Puget Sound from the train this morning

Last night I had the great pleasure of sharing a stage with Peter Levine and Jodi Sandfort at the University of Washington. We invited to come together by Jodi who is the Dean at the Evans School of Public Policy and Governance at UW and a long time friend and Art of Hosting practitioner. I have been raving about Peter’s work for a long time and she decided to covene us and about 60 others people for an evening talk looking at civic education when we are surrounded by AI.

As regular readers of this blog will know I hold Peter in high regard for his writing and ideas, easily accessible on his blog and through his books. His recent book What should we do? lays out his theory of civic life, exploring what collective action means and how to support it. I was invited to share about how we educate leaders in times of uncertainty. We talked a little about AI, but mostly our talks were about how we educate and train for collective and individual action in a time in which our collective future is uncertain and the capacities we need in the near and long term are unknown.

Peter took apart the four words of his book title, offer a learning agenda and a set of questions associated with each. His presentation was based on this map, that sets out a framework for learning and developing practitioners who wish to engage in collective action for social change. We intersected and complemented each other in a couple of ways. Peter’s questions related to the personal were a common starting point and the treatment of the SHOULD question caught my imagination.

Asking what we SHOULD do brings with it ideas, assumptions and a moral view about how action should proceed from a given moment. When there are options, the question of what should we do becomes a moral one. In a time when large forces, like AI, populism, climate change and economics are shaping the nature of uncertainty in our world, the question of how we should “host the question of should” becomes something that leaders and policy makers need to deal with. The nature of what we don’t know is changing, and the “should” becomes a precipice over which we are looking. This provides us with possibilities if leaders can avoid foreclosing the debate too soon. Leaders who are uncomfortable with uncertainty can fill the space with anxieties. An authoritarian can privilege the “DO” over the “SHOULD” and save everyone th trouble of having out the choices.

But resting in the SHOULD may be where the source of new action lies. Confronting new forms of uncertainty requires a pause before we figure out what do to. If you make sense of the world using Cynefin, this is Aporia.

After Peter spoke I shared the story of how we developed Leadership 2020, a learning program for emerging social services leaders that ran between 2009 and 2019 in British Columbia. That program was based on a curriculum that emerged from a generative framework of capacities. We identified leadership that was Personal, Practice-based, Perceptive and Participatory and crafted nine-month programs for cohorts of 25-35 people. Over the ten years we ran the program nearly 400 people joined us and learned together. The sort-of-hdden agenda of the program was to bring together a large cohort of social services leaders who coupled begin practicing this form of leadership together and be ready for the moments of collapse or crisis that are inevitable in any system.

It was a lovely evening, meeting some of the students and faculty at Evans and sharing a meal together. Now back on the best train ride on the coast, the Amtrak Cascades to Vancouver. It’s one of the few trains in North America from which it’s possible to spot a whale.

I’ll leave the text I prepared of my talk below the fold here. I departed from it on the night, but it captures well what I wanted to say.

---

What could civic education look like when we are surrounded by AI?

I’d like to begin by thanking Jodi and the Evans School for inviting me. I’ve admired Peter’s work for a long time, so it is a pleasure to finally meet him and to be here with him this evening.

I’m allergic to predictions about the future.

When my kids were young, I used to tell them that the jobs they would eventually have probably hadn’t been invented yet. I wasn’t trying to frighten them. I meant that the most useful thing they could learn was how they learned. The tools would change. The technologies would change. The work itself would change. But if they knew how to acquire new skills, how to stay curious, how to recognize when they didn’t know something, they would probably be all right.

I feel much the same way about AI.

I don’t know what AI is going to do to civic life. I can imagine it enabling some forms of participation and disabling others. I can imagine it distributing certain capabilities and concentrating others. I can imagine it changing how we organize, how we make sense together, how we encounter institutions and how institutions encounter us.

But I don’t know and that uncertainty seems to me to be the place to begin.

If we don’t know what civic life will look like in twenty years — or five years — then perhaps the educational task is not primarily to prepare people for the future we imagine. Perhaps it is to cultivate capacities that remain useful when the conditions change.

About fifteen years ago in British Columbia, we faced a version of this question in the community social services sector. A generation of leaders was approaching retirement. The Federation of Community Social Services wanted to think seriously about succession, and Jennifer Charlesworth, who was leading the Federation at the time, spent a couple of years asking people around the world a deceptively simple question:

What is the leadership your world needs now?

That question eventually became Leadership 2020, a program that brought together people from community agencies, government, and Indigenous child and family service organizations.

We became interested in a distinction my friend Harold Jarche makes between temporary skills and permanent skills, which I like much better than “hard skills” and “soft skills.”

Temporary skills are important. Budgets matter. Policy matters. Legislation matters. HR matters. Technology matters. Knowing how a bureaucracy works matters. But those skills change with the tools and the context. You can usually learn them when you need them. We used to joke that if there is a YouTube video explaining how to do it, it is probably a temporary skill.

Permanent skills are different. They are the things you never finish learning.

In Leadership 2020 we described four of them: personal, practice-based, participatory and perceptive.

Personal, because leadership always involves the person who is leading. What happens to me under pressure? What am I afraid of? What assumptions am I carrying? How is my position or my power affecting what I can see?

Practice-based, because none of this is finally mastered. You act, you notice what happens, you learn, and you try again. That sounds obvious, but it is difficult in a field like child and family services where failure can mean real harm. When failure is unthinkable, innovation quickly becomes unthinkable too. So part of the practice is learning how to experiment responsibly, how to create enough safety to learn without pretending that the stakes are low.

Participatory, because the capability to address difficult situations is distributed. Families know things. Communities know things. Front-line workers know things. Policy people know things. Leadership is not simply about having better answers. It is about creating ways for the intelligence that is already present to become consequential.

And perceptive, because complexity requires us to become better at noticing what is actually going on. What has changed? What forms of power are operating here? Where are the openings? Where are we imposing an old answer on a situation that has become something else?

Those capacities feel as relevant to civic life around AI as they did to social services successorship fifteen years ago.

AI will undoubtedly be useful. It already is. There are things machines can do faster and better than we can, and there is no particular virtue in preserving every administrative task as a human activity. Sometimes a rule is useful. Sometimes automation is useful. Sometimes the most humane thing we can do is remove unnecessary work.

But civic life is not only a problem of efficiency. It also involves judgement and responsibility and exercising legitimate authority. It requires us to negotiating conflicting values. We need to be able tp reading a situation, encounter difference and know when the rule no longer fits. So one of the capacities we may need most is discernment about what we can safely hand over and what we need to continue practising ourselves.

Sometimes relationship is the work. Sometimes judgement is the work. Sometimes participation itself builds the civic capacity we are going to need later.

Leadership 2020 taught us something else about this. Many of the people who came into the program had an activist backstory. They entered the sector because they wanted to change things. They were advocates, organizers, critics of institutions. They wanted systems to become more humane and more just. Their place in civic life was somewhere between a community activist and a steward of the system. They worked with both parents and kids and folks who wanted change, the those with the power and authority and resources to make change. Or resist it.

But these were emerging leaders and not surprisingly, some of them started being offered jobs as executive directors, CEOs and senior leaders in government. This was exactly what we had hoped for but it exposed a tension we had not fully appreciated.

It is one thing to know what is wrong with an institution. It is another thing to become responsible for one. People found themselves moving from advocacy into authority, from criticism into stewardship, from organizing into governance. Some were genuinely hesitant. “What if I screw this up? What if I become the thing I have spent my career criticizing? What if I begin using power in the ways power was once used against me?”

Those are not technical questions. But people can be prepared to live with them. They can practise self-awareness. They can get better at noticing power and context. They can develop habits of participation so that, when they acquire authority, they do not confuse authority with omniscience.

And this, I think, is one place where civic education can really matter. We often talk about civic skills as though they are mostly about conversation: listening across difference, depolarization, dialogue, participation. That matters but governance matters too. Policy, legislation,. budgets and institutions matter. If we want better institutions, somebody eventually has to become responsible for them and responsibility is not the same thing as critique.

There is a further complication. Institutions naturally reproduce themselves. Leaders tend to prepare successors who will preserve what they have built. Founders look for people who will carry forward the legacy. Organizations teach newcomers how things are done here. Some of that is necessary for good mentorship and organizational continuity but succession can also become a way of protecting an institution from the future.

If circumstances have changed, the successor’s responsibility cannot simply be to reproduce what came before. It must include the capacity to ask what still deserves to exist, what needs to change, and what has become an obstacle to the very purpose the institution was created to serve.

That seems particularly important in the age of AI.

We are rapidly weaving these technologies into the infrastructure of our lives. We are using them because they are useful. We may come to depend on tools that we will later need to regulate, constrain or redesign. nIt is difficult to challenge the things that serve us. In many ways.

So perhaps the civic capacities we need are not principally about predicting the effects of AI. I think we can reliably say that a significant technology like AI crates previously unknown uncertainties. These uncertainties demand that we remain capable of noticing our dependencies, noticing how power is shifting, noticing when convenience has become a constraint, and being willing to act on what we see.

At the beginning of Leadership 2020, we used to say that we had a secret agenda. While we were very devoted to their learning and development, we had a bigger purpose in mind. We told participants that systems periodically crack and reorganize. About every 30 years, that’s what happens to child and family services. The last major change in BC was in 1996. We may be on the verge of another one now.

We could not tell them when that change would happen or how it would happen or what the next system should look like.

Our hope was simply that, when an opening appeared and things came apart, that people from this program would find each other and run toward the centre, recognizing that this was the moment they had been preparing for.

Not because they had the answer, but because they had practised entering uncertainty, gathering others together to observe and make sense of what was happening, using power with some care, and taking responsibility for what came next. Over the course of their learning they had met others from elsewhere in the system that they could work with across difference, to build something new.

That is probably the civic learning agenda that interests me most in the age of AI. We can prepare people to meet futures we cannot yet know. And we can do it in a way that builds sustained relationship and commitment between them so they stay connected across differences and find each other when they are needed/

If Peter’s question is, “What should we do?”, mine is perhaps one step earlier:

What should we be practising now so that, when the moment comes, we are actually capable of doing it?