Co-Designing AI Like at Grandma's Lunch: The Generativa Campus

As I return to Rome after the wonderful experience of the Campus of Generative In Naples I can't resist, I take the PC and start writing (then it took me a week to finish writing it, but that's another story).

Generativa was a training experience on AI for culture and art, lasting several months and organized by a diverse and competent partnership with Sineglossa, Oriental University of Naples, TLON, Matera Hub And Creative BasilicataAt the end of the online lessons, we, as Ecosistemica, were asked to organize two days of co-design with the 60 people who had attended the courses, with the aim of creating communities of practice.
In this article, I will recount the experience from both a design process perspective, but also, and above all, from a relational and human perspective, bringing back what has remained alive in me: the perspective of the facilitator who is also a person among others and weaves relationships.

Genuine design

But how much design work does it take to build a context of genuine authenticity where people can get to know each other, exchange knowledge, learn, and work together?

It seems like a contradiction in terms. To have an experience that allows people to be authentic to themselves and others, a lot of design work is needed. This is because we are accustomed to hierarchical (it's not a dirty word and the hierarchy needs to be explicitly stated) command-and-control contexts, perhaps because places of knowledge serendipitous or spontaneous with other people are gradually running out (or are simply strongly discouraged by the public and political system in force), perhaps because time is becoming increasingly optimized and mediated, perhaps because we give so much importance to output and technique, that the process and the relationship are put into the background.

As I write, a question arises: how difficult is it to replicate Grandma's lunch? Grandma's lunch will never return in the form it once was. Grandma was a social, gender, and narrative construct of postwar Italy, brought about by a fortuitous series of phenomena: people lived longer than in previous eras, had adequate pensions, and could care for others. Grandma cannot be replicated, but we can learn a lot from this idea. Grandma as an individual is meaningless if she isn't inserted into a relational context. Her dish isn't gastronomic perfection, her home isn't the most luxurious, the family she lives in isn't ideal. That set of intertwined relationships—sociality, affection, local and seasonal products, colors, memories—acquires an inestimable value. Grandma hosts, creating the conditions so that those invited have nothing to think about other than enjoying the experience. The experience isn't aimed at the output of eating that particular dish. Grandma's dish cannot be replicated, it cannot be forcibly constructed in a laboratory, because it would lack its relationships.

We at Ecosistemica learned from Grandma. In Naples, we wanted to be good hosts before thinking about technology. We chose to nurture and enable a celebratory, non-hierarchical, non-competitive environment that values the individual in their complexity before the professional. We chose to design a co-design workshop for products, services, and artificial intelligence systems, giving as much space as possible to relational intelligence—not that of the individual, but that which emerges from people's connections, connections with a process, with tools, with a place and time, situating what Levy called collective intelligenceWe chose to do it in the simplest way possible, without frills but guided by shared postures, taking inspiration from Design Jam, from theArt of Hosting and using tools that have existed in design for decades and that were accessible to everyone. 

The open design

When I started working 15 years ago, free, accessible, and locally distributed co-design events were the order of the day. Design Jams, for example, involved designing systems and products to improve people's lives, platforms for environmental sustainability, and services to support democracy. All of this happened in groups with strangers. I remember a flourishing of accessible tools for working together, canvases for designing anything, downloadable from an internet that wasn't yet drastically compartmentalized into large, closed, proprietary platforms, where the goal wasn't to maximize profit with every interaction, and where the traffic generated by bots It was still a minority. The internet was much more similar to its early days, a frontier of exploration, but above all a collective project where each person contributed in their own way. Fifteen years ago, I could download all the design tools I still work with for free, because someone was experimenting with them, modifying them, and releasing them under an open license. While organizing the Campus, I had a hard time finding open tools and databases, so I took the ones I had under Creative Commons and redesigned them.

At the end of the Campus, more than one person asked me if we would share the tools (canvases) we used. The answer seemed implicit: "Of course!" The tools are there to help us reflect. I believe it's essential to share our creations, and I believe it's essential to share what allows us to reflect together.

Another participant confessed to me that she had a wonderful idea for an AI but didn't want to share it with me for fear I'd copy it. This got me thinking. From a young age, we're taught not to copy; it's wrong, it's deplorable, it's disrespectful to the person you're copying from, to the supervisor, and to yourself. If copying means suppressing your critical thinking, then I can agree, as in the case where copying significantly harms the person you're copying from. But sometimes copying means allowing an idea to have greater resonance, benefiting more people, allowing you to skip the more boring stages and get straight to the interesting ones. In some cases, you can even grant full copying rights; this is usually done with Creative CommonsImagine what a world it would have been if the penicillin had been patentedImagine, too, what a world it would be if the rights to use and reproduce all those cultural, artistic, and technical products that are now closed and represent a potential resource for humanity were accessible. On this topic, Ippolita's reflection on Open is not Free, which I join because I feel, as a thinker and creator, the strong responsibility and pleasure of sharing.

In our case, we are trying and will continue to try to make all the tools we build open and free. For example, for this occasion, we have uploaded the Actors' Map, the Wheel of Privilege, and the Journey Map used during the Campus on the Zotero where we usually archive the slides and materials from the Ecosystemic workshops and training courses.

Relational artificial intelligences

An article about an artificial intelligence co-design workshop that only addresses the topic of AI at the end? Yes, I strongly believe that technology should be addressed at least together with the issues bio-anthropo-psycho-socio-historical and not as a hype to be capitalized at all costs, not as a strategy for defining self-fulfilling propheciesI think it's important to talk about access to tools, about ownership, of power, of humans and of relationships. I believe it is important first of all to define our collective objectives and then build the technical conditions to achieve them. It is impossible not to refer to Salvatore Iaconesi when talking about these topics:

For humans, AI is an **EXISTENTIAL** issue, which has psychological, cultural, social, political, relational, economic implications and concerns fundamental human rights.

More than a technical campus, Generativa was a laboratory for shared reflection on where we humans wanted to go, a laboratory of possible directions, needs, desires, and conflicts that perhaps, until now, had remained closed and fragmented among all the people who generously participated.

The outputs were multiple:

• AI as a support for discreet sexual-affective education;
• AI to connect archives and make searching easier;
• AI to create music-based relational environments;
• AI for territorial social research;
• AI to encourage younger generations to read more and study better;
• AI as a colleague to co-design human-human and human-nonhuman agent work policies;
• AI for the personalization of museum itineraries;
• AI to stimulate intergenerational development in areas at risk of depopulation;
• AI for cultural matching of people doing research;
• AI to encourage residence in inland areas.

I won't go into detail describing or commenting on all the ideas; perhaps my facilitation colleagues, Stefano Capezzuto, on the AI ethics side, and Daniele Gambetta, on the AI technical feasibility side, will do that. I hope the projects that emerged will be shared soon, because I believe they have provided truly significant insights that deserve to be shared, implemented, and, in some ways, even copied.