“In thirty years working in museums, this is the first time I’ve been able to make so much tailored information available to the public in such a short amount of time.”
- Boris Wastiau, Director of the Alimentarium, on the launch of Alima with Ask Mona.

‍

In most museums, visitors arrive knowing almost nothing about what they are looking at. They rely on wall texts, guides, and audio tours. The institution knows; the visitor listens.
‍

At the Alimentarium, the world’s largest museum dedicated to food, things work differently. Everyone eats. Everyone grows up with tastes, habits, and memories tied to food. When visitors enter the SYSTEMA ALIMENTARIUM exhibition, they already come with opinions. About fermentation, edible insects, or what their generation really eats. This is not the posture of someone standing ignorantly before a masterpiece. It is a conversation waiting to happen.
‍

Deploying a conversational agent in this context changes the nature of mediation itself. Alima does not broadcast information. She engages in dialogue with people who already have something to say.
‍

When the Format Meets the Subject Matter

The data confirms it. One month after launch, on April 17, 2026, the average session length reached four minutes, with visitors exchanging an average of 8.4 messages per conversation. The logs do not reveal isolated factual questions. They reveal long, fully developed discussions.
‍

One visitor engages in six consecutive exchanges about the microbiology of lacto-fermented dry sausage, not as a student, but as someone challenging the information. Another questions Alima, the name of our conversational agent, about the exact number of fake dishes in a crate because he counted more himself. A group uses her as an educational referee in the middle of a debate: “That’s valid, right, Karine?” Someone else asks why the Magnum photography selection does not include Nestlé factories.
‍

These exchanges would never happen in front of an Old Master painting or an archaeological artifact. They happen because the subject is food, and every visitor approaches it with personal experience that seeks confrontation, confirmation, or contradiction.
‍

“In thirty years working in museums, this is the first time I’ve been able to make so much tailored information available to the public in such a short amount of time,”
said Boris Wastiau, Director of the Alimentarium, at the launch event.
‍

Wherever the Visitor Looks, Alima Is Already There

The installation includes twenty distinct agents, deployed just four months after the project officially began in December 2025. Each one is tied to a specific area of the exhibition: the Galaxy of Taste, the Critical Zone, the room titled Raw, Cooked, and Rotten, or The Fork Planted in Lake Geneva. Every agent speaks from within the space it inhabits. When a visitor scans a QR code in front of an object, the agent already knows what they are looking at.
‍

All agents respond in five languages using a knowledge base created and validated by the museum team, with content fully controlled by the institution. This choice matters. The institution controls what Alima knows, and therefore what she says.
‍

The agent has a first name. Alima borrows hers from classical Arabic, somewhere between knowledge and wisdom. She embodies an identity designed for 15 to 25-year-olds, an audience museums often struggle to attract before the end of their studies. Giving a conversational agent a first name, a voice, and a posture means shaping the relationship visitors will have with it before they even ask their first question.
‍

“AI is here to serve curiosity. The whole system depends on visitors asking questions based on what they see. They become active participants in the experience.”
Marion Carré, President of Ask Mona, summarizes the philosophy behind the project this way.
‍

What the Museum Learns in Return

The system produces something no mediation panel ever could: a trace of what visitors actually wonder about.
‍

Every conversation leaves a record accessible to the curatorial team. When an answer proves incomplete, the knowledge base can be enriched. A visitor standing in front of a flame atomic absorption spectrophotometer asks a question. Alima can place the instrument within the history of twentieth-century food analysis techniques, but does not yet have a precise definition available. The team identifies the gap, adds the missing content to the knowledge base, and future visitors receive a complete answer.
‍

Attendance figures cannot generate this kind of signal. Conversations can.

What the Alimentarium Foreshadows

The Alimentarium is the first Swiss institution to deploy this kind of architecture. It is also, to our knowledge, the first museum ever to receive a question like this from a visitor: why does the Magnum photography selection not show Nestlé factories? No wall text would ever have triggered that exchange.
‍

For decades, cultural mediation has relied on a top-down logic: the institution produces knowledge, and the visitor receives it. Wall texts, audio guides, labels, all are built on that asymmetry.
‍

The Alimentarium is the first Swiss institution proposing something different. With Alima, the visitor becomes an interlocutor. They arrive with questions, opinions, and lived experience. And the institution responds on equal footing.

That is as much a shift in posture as it is in technology.

‍