Google Futures Lab | Students build AI learning prototypes
Google has highlighted AI learning prototypes built by students in the Google-funded Futures Lab, a partnership with the University of Waterloo. The projects include Kanji Garden for Japanese learning, SignFluent for real-time American Sign Language feedback, and MuscleMemory for AI-assisted exercise form guidance, showing how students are testing practical education tools with AI and UX prototyping.
Google shows how students are prototyping AI tools for learning
Google's article focuses on Futures Lab, an eight-week intensive AI and user experience prototyping workshop created through a partnership with the University of Waterloo. The program gives students from fields such as computer science, business, and natural sciences a space to design working prototypes around the future of learning.
For designers, the relevant point is the workflow. These projects are not only technology demos; they combine AI, user research, accessibility, feedback loops, and interface design to test how learning tools could become more adaptive, visual, and responsive to individual users.
What students built in the Futures Lab
Kanji Garden is a language-learning prototype that teaches Japanese through immersive AI-generated stories and visuals instead of relying only on rote memorization. The idea is to make abstract language learning more contextual, visual, and emotionally memorable.
SignFluent focuses on accessibility and real-time feedback. The tool helps users practice American Sign Language by analyzing their form and giving instant guidance, showing how AI can support learning when movement, body position, and visual accuracy are central to the skill.
New workflow lessons for AI education design
MuscleMemory applies AI camera tracking to calisthenics training, offering real-time audio feedback on exercise form. While it is not a traditional classroom tool, the project shows how learning can extend into physical practice, where immediate correction may help reduce mistakes and prevent injuries.
The broader design lesson is that AI learning tools need more than model output. They need the right interaction pattern: visual storytelling for language, accessibility-aware feedback for sign language, and audio guidance for movement-based learning.
For product and UX teams, these prototypes show why early testing matters. Students used the lab to explore user-centered design, applied communication, accessibility, and product prototyping, which are all essential when AI tools are meant to teach real skills rather than simply generate content.
Availability and education context
Google presents the projects as working prototypes from Futures Lab rather than finished commercial products. The lab is led by Dr. Edith Law, Google Chair in the Future of Work and Learning, and is designed to help students co-create technology for the future of education and work.
For designers and education teams, the practical takeaway is to treat AI learning tools as experience systems. The strongest results will likely come from combining AI generation, feedback, accessibility, clear UX flows, and careful review by educators or domain specialists before moving toward public deployment.
Daisuki's Take: What This Means for Designers
We see these Google Futures Lab prototypes as useful because they show AI design beyond simple content generation. The projects focus on learning experiences where interaction, feedback, accessibility, and user context matter as much as the model itself. For designers, that is a stronger direction than treating AI as only a tool for faster text, image, or video output.
The strongest lesson is that good AI education tools need the right interface pattern for each skill. Japanese learning can benefit from visual storytelling, sign language practice needs precise movement feedback, and exercise guidance depends on timing, body position, and clear audio instructions. That means design decisions should be shaped by the learning goal, not only by what the AI model can generate.
The limitation is that prototypes are not the same as production-ready learning tools. Before reaching real users at scale, these systems would need stronger testing, accessibility validation, privacy review, educator input, and safeguards around feedback accuracy. Used carefully, this kind of work can help designers think about AI as part of a complete learning experience, not just as an automated content layer.
Sources and Recommended Links
- Students prototype learning tools with AI at University of Waterloo | Google Blog (Official)
- University of Waterloo | University of Waterloo (Official)
- Google AI | Google Blog (Official)
