Localism in Action

Getting things done in Sussex, and why that's relevant to AI and edtech

A couple of weeks ago, walking near my home in Sussex, about an hour south of London, I had a light bulb moment about localism.

I was on my way back from our local town hall, East Court, on a bright and sunny Spring day.

I visited because I am leading an effort to reduce the speed limit on a stretch of residential road near my home from 30 mph (with many cars speeding faster) to 20 mph. I already had support from the County Council (representing the region), and needed support from the Town Council too. Rather than emailing or calling, I decided to walk (on my legs!) over to East Court to find out who would be the right person to help.

At East Court, not only did I identify the right person (the Town Clerk), but she was right there. We had a great conversation, and agreed—on the spot—that the issue could be put on the agenda of the next meeting of the Town Council. Mission more than accomplished! I headed back home.

On the walk back home in the sunshine, birds were singing and Spring was springing. And in this moment of natural enjoyment, I had my light bulb moment. It was simply this:

The concept of localism is both powerful and general.

I have been writing a lot about localism in AI and edtech, saying things like this:

The promising way forward for AI, beyond LLMs, is one focusing on purpose-fit tools that use a variety of computational methods and focus on local conditions.

and this:

[R]esponsible AI … requires a detailed focus on local conditions. These local conditions include many features such as specific applications, types of users, training and inference data, AI models and other technical details, locations, and various other factors.

In short: “Think global, act local”—a 1970s slogan that I quoted in the latter blog.

Local political action is conceptually close to local approaches to AI and edtech. The application of localism across diverse conceptual domains is what I mean by writing that localism is general.

My decision to walk over to East Court—in the hope that I would accomplish more than on the phone or email—turned out very well. That outcome wasn’t guaranteed, and it would have been different if, for example, the Town Clerk had been out for lunch. But engagement with local conditions did put me in a much better position to succeed. And the experience was enjoyable to me as a human.

When developing AI and edtech solutions, we should take ‘local walks’ to see what is going on, including human dimensions. For the time being, an AI cannot walk across town and have a great conversation with the Town Clerk. This may change over time, but the human dimension of locality will continue to be crucial well into the future.

Failure to act local can have dramatic consequences. For example, it appears likely that the Trump administration designed its dramatic recent tariff policy based on a formula recommended by an LLM. This reverses the “Think global, act local” ethos, and dramatically departs from localism by experimenting with the complex and diverse world economy using a single, simplistic formula. It remains to be seen how this will work out, but initial indications are not good.

Think global, act local, my friends.