Exploring AI Problem Formulation with Children via Teachable Machines

TL;DR

This study investigates how children aged 8–13 can formulate AI problems through participatory design using teachable machines. By pairing children with university-based adult collaborators and employing a structured Big Paper storyboard with problem-reduction prompts, the authors explore what constitutes children’s AI problems, the design metaphors they favor, and the values embedded in their solutions. Key findings show that children draw on personal life experiences, anticipate errors and recovery strategies, and predominantly favor human-centered, controllable design metaphors (e.g., Supertools, Control Centers) while expressing instrumental values like capability and responsibility and terminal values such as family security and a comfortable life. The work offers practical guidance for designing future participatory AI activities, advancing AI literacies for youth, and informing UX practice through design metaphors and value-centered analysis.

Abstract

Emphasizing problem formulation in AI literacy activities with children is vital, yet we lack empirical studies on their structure and affordances. We propose that participatory design involving teachable machines facilitates problem formulation activities. To test this, we integrated problem reduction heuristics into storyboarding and invited a university-based intergenerational design team of 10 children (ages 8-13) and 9 adults to co-design a teachable machine. We find that children draw from personal experiences when formulating AI problems; they assume voice and video capabilities, explore diverse machine learning approaches, and plan for error handling. Their ideas promote human involvement in AI, though some are drawn to more autonomous systems. Their designs prioritize values like capability, logic, helpfulness, responsibility, and obedience, and a preference for a comfortable life, family security, inner harmony, and excitement as end-states. We conclude by discussing how these results can inform the design of future participatory AI activities.

Figures (4)

• Figure 1: Our co-design study comprises: a) modified Big Paper with a structured storyboard that b) child-adult pairs use to frame and discuss their ideas and then c) present their teachable machines while a researcher summarizes their input.
• Figure 2: Our work is part of multiple co-design sessions. In session 1, technology immersion druin1999cooperative, children engaged with Google’s Teachable Machines, trained the classifier by exerting control over the input (training examples of the objects to be recognized), and tested each other’s training efforts. In session 2, also a technology immersion, children engaged with a teachable augmented reality application in a museum. The application, developed by our team, enabled children to control both the input (training examples of objects in a museum exhibit) and the output (their own 3D designs that were triggered upon successful recognition of the object). In this study, we present results from session 3, that engaged children in AI problem formulation via a modified "Big Paper" storyboarding activity. In session 3, children explored how they might formulate and approach a machine teaching problem of their own design, influenced by their every day life experience.
• Figure 3: Children's storyboards: a) Brian's smart-glasses based "Problem Machine" helps solve any math problem just by looking at it, b) Ed's robot helps you with chores such as folding clothes, c) Luke's "Robocalculator" can travel to you and solve math problems, d) Nancy's portable food store makes food for you if you get the ingredients and tell it the recipe, e) Ollie's smart light controller, "Light", recognizes her and turns on a light when she wakes up, f) Adrian's robot security guard, "Handy Helper," keeps his younger brother in check and protects him from bullies, g) Alan's wall-mounted emotion detection machine, "Mr. Crab," responds to your emotions, h) Denny's smart light controller, "Claude," creates your ideal homework environment, i) Penny's safety bracelet for cheerleaders alerts you if you are about to fall and blows up a mat to cushion any fall, and j) Kevin's holographic smart home security system, "Holo," alerts you of possible intruders and open windows/doors.
• Figure 4: These Sankey diagrams illustrate the relations between the design metaphors and human values in the children's designs. Each flow line represent the strength of a link (or affinity) between the two e.g., "Active Appliances" are closer to Instrumental Values like "Helpfulness," "Capability", "Logic" and "Responsibilty"; and the Terminal Values of "A Comfortable Life" and "Family Security."