The AI Chef Trainer is an educational web app that introduces children to the role of data in machine learning (ML) through the engaging task of recipe recommendation. Initially, students tested the AI Chef's capabilities by selecting from a list of ingredients to see what the system recommended as possible recipes. After observing the recommendations, they contributed by adding their own recipes—each being a set of ingredients and a corresponding recipe-name—which were used to retrain the model and finally re-tested recipe suggestions. This cyclical process of testing, contributing, retraining, and post-training testing provided students with hands-on experience in how AI systems learn and adapt over time based on new data. We tested our software with middle school students. The results indicated that students recognized the importance of both data quantity and specificity in the training process. 45 of 52 students entered recipes, and 26 of the 52 tested their own recipes using the specific ingredients they entered. Students were introduced to the concept of confidence percentages via the AI recipe suggestions. Even as the primary focus was the role of data in machine learning, the AI Chef Trainer software also served as a window into students' cultural expression and personal preferences.

This paper explores the evolution of artificial intelligence (AI) and machine learning (ML) educational tools and curricula for children, describing a spectrum from playful exploration to structured pedagogy. We address three research questions: (RQ1) What themes are revealed in the literature on AI/ML for children? (RQ2) In what ways do the development of AI/ML tools and curricula co-evolve? (RQ3) How does the historical context of the field frame current work on AI/ML for children? We observe a co-evolution of tools and curricula, where researchers build new tools and curriculum developers adopt tools created by others. As tools prove successful, they become widely available and are adopted by a broad range of users. Additionally, we note instances where tool creators take on the role of curriculum developer. Recognizing the need for a clear overview of AI/ML educational resources, we introduce major themes of tools, tool-forward, and curriculum-forward works. These themes are designed to help educators, researchers, and curriculum designers in discerning the fundamental nature and potential applications of resources. By distinguishing tools that are standalone from those integrated into curricula, our framework supports strategic decisions in resource adoption and development. Drawing on the field's history, we outline a continuum of tools and platforms, from constructionist elements to interactive play environments.

Artificial intelligence (AI) and machine learning (ML) have a deepening impact in our world. For empowered citizenship and career readiness, elementary and middle school students need to understand these technologies. This poster reports on five original interactive AI and ML software exhibits tested by 125 elementary and middle school students aged 7 to 14 years. Four themes emerged: Students recognized that AI and ML systems can process data from cameras (perception); they saw that these systems responded to their training input (trust); they appreciated the practical import of AI/ML systems (affective and cognitive attitudes); and students were introduced to models and modes (specialization).