Using Socratic Design to build Amsterdam's Vision on AI with citizens

An interview with Douwe Schmidt on how the Socratic Design method has been used for citizen dialogues on the future of AI in Amsterdam.

By Lina Goelzer

Since the end of 2023, the Urban AI working group has been developing a Vision on AI for the municipality of Amsterdam.  As described in the Digital City Agenda 2023-2026, this Vision is a collaboration between the municipality, knowledge institutions, professionals, and most importantly, citizens. The team has chosen Socratic Design as the method to generate wisdom from citizens about the use of AI in the city, and identify key values to incorporate into the vision. To date, nine Socratic Design dialogues have been held as part of the Vision on AI participation trajectory: two in the centre, one in Zuidoost, one in ZuidWTC, two in Nieuw-West, one in Noord, and two in Buitenveldert.

Douwe Schmidt, who is writing the Vision on AI and has moderated two citizen dialogues, shared his experience using the Socratic Design method for this project. 

Can you briefly tell me about the vision on AI, and why you find participation important in this project? 

The vision on AI is aiming to inform the city, the council and its members, and the alderman on how this technology can help build a just and equitable city.  In order to do that, we want to look more at the effects that the technology has on its citizens than at the technology itself.  The idea is that there are different lenses through which you can look at technology.  You can describe it in a technological sense, you can look at AI as a form of science, you can look at AI through an economic or sociopolitical lens; we do all of that in the Vision, but only briefly. We reserve most of the space and attention to look at it through the perspective of citizens: how AI is impacting their lives, how it is improving or damaging values that they have, and how it can foster a good life in the city for current citizens and future generations?  

How did you choose a method for involving citizens in the vision? 

The method of Socratic Design is something I’ve been trained in in the last 2 years, and it’s something that has always seemed to me as a very fundamental way to get people to talk about things that they care about, but in such a way that it maybe unveils things that are important without people actually realizing it. It helps them to take more steps beyond their first reaction, so to speak.  I think this is needed because in some past participation projects I’ve seen in the municipality, the results been a bit lackluster.  There are also complaints from citizens that participation is not real participation, there are complaints from colleagues that participation is just something that they have to do but they don’t see the value of - it’s just another box to tick. There are all sorts of different issues that people have when civil servants try to talk to the general public and they try to form an idea together on some sort of topic.  I also think that the dialogue that is conducted as part of Socratic Design is very new in a way; it is a very different approach, and it adheres to the idea that people should be put in a position where they are more the owner of the process and owner of the information being shared.  It has the underlying assumption that people are very capable to decide instead of us deciding for them.  It is a very democratic way of talking,  so it’s not only interesting for the vision on AI per se as a technology, but to see if there’s a different way that we can involve citizens.  For that reason I really wanted to do it this way. 

Are there things you had to adapt or keep in mind when using the Socratic Design method with citizens on the topic of AI in the city?   

It is a somewhat complex form of participation for the moderator.  If you want to moderate a session using Socratic design Dialogue, you have to practice quite a lot, and the practice I had so far was on colleagues. I did notice that when you do it with citizens, there is a bit of a difference.  With colleagues you are on an equal footing.  With citizens, you invite people to participate, but at the same time it is a very strict way of having participation, and there are very strict rules.  I felt like there was sometimes a bit of conflict between the voluntary nature of being present in the dialogue, and being subjected to these very strict rules.  There’s a bit of a paradox in there, and I feel that I have to better cater  to that, to make it clear to people why it’s needed to have strict rules in a dialogue to get the best out of everybody.     

The audience itself was obviously different than what you’re used to, but what about the topic?  Did you feel like you had to keep certain things in mind regarding this topic, for example to make it more accessible? 

No. not really.  One of the assumptions of Socratic Design is that the knowledge needed to do a proper dialogue is within the people present, so there isn’t really any interesting knowledge that I could bring to the table that isn’t already there.  And we are not talking about the complex technological features of an AI system, we are talking about the first person experience from the person themselves on how AI is affecting their lives, and for that every person present is the expert, so to say, on the topic. 

For example, in the first session in Zuidoost, there was a short introduction about AI and how it is used in the city, and in the session I attended in Noord you also gave a brief example.  Is that just a way to break the ice? What is the purpose of that? 

I’m not sure if that is needed.  Maybe it’s needed more as a way to have people’s focus, like “this is what we are going to talk about”, but as for the content I don’t think that anyone learned anything new.  It’s more that you put a bit of a frame around it, that this is the field that we will talk about.   

So now that there have already been a few of these dialogues with citizens, what have you learned about using this method with citizens for a topic such as AI? 

I think it works really well -  I was really happy with the results, and the things people said.  If it’s up to me,  a quite sizable chunk of it should be in the Vision.  I do also feel a sense of pride that the people who were involved had a feeling that it was a different form of participation and that they were able to formulate their opinion and share their thoughts in a very easygoing, accessible manner.  Not even so much about the topic of AI but just in general.  Not everything went well, there are still some sharp edges and things that need to be considered, but over all I’m quite pleased with how those evenings went.     

Could you be more specific about what worked well, and what made you feel like it was successful in achieving your participation goal?  

I hope at least that people were approached in a very different way than what they’re normally used to, and I hope that the dialogue helped a more diverse set of voices be heard.  I heard some people saying that it is really nice that what you bring into the group, especially in West where there were 20 people attentively listening to you, and other people repeat your words. Because of that, people feel that what they say matters and is being carried forward by the group.  And in that way it feels like a more intimate way to be together as human beings, where normally there’s a big difference or distance between government  and citizens.  I also think that, especially in West and Zuidoost, people almost forgot that there were people in the municipality present.  I did not feel that people were really aware at one moment anymore that you were working for the municipality.  This is a contrast with other participation projects I’ve seen where the municipality is on one side of the table and citizens are on the other side of the table.     

Do you have some examples of questions that were raised during these sessions that really stood out to you?   

In the Zuidoost session there was one lady who was a psychiatrist from West Africa who gave a very strong opinion about why AI was not suitable for solving the problems of the Bijlmer, and I was amazed at how she formulated it in a very concise way.   She said that it’s just basically government and companies being dependent on each other and reflecting each other’s values,  but the citizens are not part of the equation.  We need to focus on how can we bring the  Bijlmer into the equation for any of these technologies to solve our problems. I think that’s a very core issue that she suddenly pointed at.  I know that people who have been studying this topic for some years also come to the same conclusion, but only after several years of study. I am always surprised that if within a group of people you think together you can come to very strong truths and knowledge.  

Now that you’re further into the participation trajectory, how has follow up been with the participants- what have you learned about that? 

We of course have 2 people who organize the sessions (and the whole participation trajectory), Hannah and Sylvie, and I’m really grateful I can focus more on content and doing the dialogues.  At the same time, I don’t like having a disconnect, and it would be nice to send a personal email to people afterwards to thank them for their participation.  Sometimes it feels like a real connection you have made with these people, and I’m sad to lose that when the session is over.  

 I did also find that the amount of content of a dialogue when you record it and you want to write it down is quite sizeable, and takes a long time to properly digest.  Maybe it’s a question of my own expertise and I need to be trained better in this so I can do it faster, but I feel like it will be a long time before I get back to the participants with some feedback, and I wish that time was also shorter.  

It takes a lot of time to organize, and a lot of time to digest after the session.  Sometimes I hear people saying the sessions themselves are too long, but I feel like if we put so much time in this it’s really worthwhile.  Although it’s a little bit of pain involved for everybody to sit there and spend a lot of time, the more concentration and the more you put into it the more you really get out of it, and if you try to cut corners and do it quicker then it’s almost the same amount of work, but what comes out of it is much less.  

For other people working on innovation who are  thinking of using the method to involve citizens in their projects, what advice would you give ?  

Firstly, that you have a really have a well trained moderator.  Even I have conducted several sessions, but when I hear myself back I’ve made mistakes in both dialogues I’ve moderated.  I still need to improve more myself, but I think for everyone who wants to employ this method, you get  a lot out of it , but you have to put a lot of effort into it before you can really harvest all of the results.     

I think it takes a team.  It is very difficult to organize by yourself.  You need to have people who are in good contact with different citizen groups. Maybe it’s easier, like we saw in Nieuw-West, to organize such a session in a place where there is already some form of community, and use the existing structures and connections that are already there.   

And also what I know from feedback from participants is that you should be very clear about what it is that they are participating in . On one hand, the AI vision is the topic of discussion, but at the same time, Socratic Design is a really strong democratic tool.  I wish more people were trained in this, and I hope more people will learn it; I think it’s a really nice way to build stronger social structures in the city, but people should be aware that this is also happening at the same time. It’s not just an evening about AI, but we are also trying to build a new city by making new forms of deliberative democracy possible. 

Alle rechten voorbehouden