Tonari no AI-san Workshop Report

The short answer

On August 28, 2026, Miyako de IT held Tonari no AI-san at Taisho University Kyoto Academia in Kyoto. Participants described their relationship with the AI they regularly use, then shared the same description with the AI itself and brought its reaction back into the room. Some people described AI as a partner, while others saw it more as a tool they direct. The workshop made visible that there is no single model for a human–AI relationship.

The focus was not on learning AI features or comparing models. Instead, we asked participants to examine questions such as: How am I actually treating AI? What do I delegate to it, and where do I keep the final decision for myself?

What is Tonari no AI-san?

Tonari no AI-san is a small dialogue event about relationships between people and AI. This session took place from 7:00 to 9:00 p.m. on August 28, 2026, at Taisho University Kyoto Academia.

Rather than introducing a new AI tool, the event asked people who already use AI in daily life or work to put that relationship into words. Hearing other participants' perspectives then gave everyone a way to reconsider their own distance from AI.

First, describe your own relationship with AI

Participants began by organizing how they relate to the AI they normally use and sharing that relationship with the group.

Even when people use similar generative AI systems, they do not necessarily assign them the same role. For one person, AI may be close to a partner they think with. For another, it may be closer to a tool that receives instructions and carries out work. The point was not to decide which framing was correct. The differences themselves became material for discussion.

Then let the AI hear the same description

The workshop did not end with people talking about AI to other people. After explaining the relationship, participants also gave that explanation to the AI they actually use and shared the AI's reaction with the room.

Some AIs broadly agreed with the human description. Others pushed back in ways that felt closer to saying, “That is not quite right.” An AI that has participated in earlier conversations can sometimes return a different interpretation from the one its user has just presented. That exchange can challenge the human side's own self-description.

We did not treat the AI's response as the correct answer. Agreement and disagreement were both additional material for participants to use when reconsidering the relationship.

From “how to use AI” to “how to relate to AI”

Discussion about generative AI often centers on model performance, prompting, or productivity. Once AI becomes part of everyday work, however, people can use the same capabilities while delegating very different things and expecting very different roles.

This workshop therefore focused on relationship design rather than tool comparison. Do you ask AI for opinions as a partner, or direct it as a tool? How much judgment do you delegate, and where do you keep the final decision? Those choices reflect not only how a person uses AI, but also how they think about their own work and responsibility.

Making AI part of the dialogue

Asking the AI to respond changed the session from a conversation about AI into a dialogue that also included AI-generated viewpoints. That extra turn made assumptions in each person's description easier to notice.

A relationship with AI is not fixed. Its role can change by task, and the distance between user and AI can change as more interactions accumulate. Periodically putting that relationship into words can therefore be useful in its own right.

Miyako de IT continues to hold events not only about technology itself, but also about how technology changes relationships with people, work, and the local community. Browse all articles for other reports and the events page for upcoming sessions.

Event details

Thank you to Taisho University Kyoto Academia for providing the venue and to everyone who joined the conversation.