AI x Lastinger Center
We believe that purpose-built AI has the potential to enhance human capacity to draw on themes, make recommendations, and impact student learning outcomes.
What is Purpose-Built AI?
Purpose-built AI are technologies designed for a specific learning goal or problem, in service of enhancing learning, in partnership with students, teachers, and families. In education, AI should align to a vision for learning and apply modern learning principles (US Department of Education, 2023). We apply these principles in our AI initiatives to improve the quality and scale of the services we offer. Here is what we have been working on:
Vollee
In our professional learning system, we implemented technology from Vollee that acts as an automated assessor, providing personalized feedback while instructors serve as human-in-the-loop to ensure quality. This hybrid approach allows learners to work at their own pace and practice as often as they want while benefiting from explainable AI evaluations. The system significantly improves the efficiency and accuracy of human assessors. Vollee’s solution enables faster, more consistent feedback for learners, who benefit from quicker assessments and shorter wait times for personalized feedback, allowing them to implement instructional improvements without delay.
Kibeam
The UF Lastinger Center for Learning and Kibeam Learning, Inc., have partnered to create the screenless Kibeam Wand™, a small hand-held device designed for young readers to enhance reader engagement with books and inspire fun and interactive family reading experiences. Using AI, the Wand™ reads, checks understanding and encourages children to tilt, point and swirl the device as they engage with a paper book.
ALTER-Math
We are partnering with researchers at the University of Florida, Vanderbilt University, Stanford University, and Duke University to develop an AI-driven chat-based feature to improve mathematics skills. The ALTER-Math avatar uses teach-to-learn principles in which students teach the avatar when it demonstrates mathematical misconceptions. The teachable agent helps teachers to identify where they can focus with students on specific concepts.
Data Lagoon
The Lastinger Data Lagoon aims to accelerate research and development in education, including AI development, by sharing data without barriers. This datahub contains large scale datasets from professional learning programs in early learning, literacy, and mathematics, as well as 6-12th grade math students through programs provided by the University of Florida Lastinger Center for Learning. Our data enable a rich view of education in Florida, extending currently to hundreds of thousands of participants and billions of records over almost ten years.
What concerns are we keeping in mind?
Privacy and Data Security
Privacy and data security are a top priority. Our work maintains the rigorous privacy standards established by the Family Educational Rights and Privacy Act (FERPA).
Biases, Plagiarism and Misinformation
Considering limited, inaccurate, outdated data-sets are likely to exist, our work will continue to be maintained by researchers, faculty and staff who are experts in their fields.
What’s next?
By following the models and recommendations of trusted sources, the Lastinger Center is exploring use cases of purpose-built AI tools to better understand the viability of its use and determine appropriate, controlled environments in which we can test, learn and innovate.
To learn more about projects with new and emerging technologies, visit the Learning Lab.