AI has rapidly shifted the education landscape, and at VitalSource, we believe using AI as a tool to create impactful, scalable learning experiences grounded in learning science research is the way forward. AI can expand the reach of proven learning methods to millions of learners. With our experience researching learning science, we have a deep understanding of the responsibility to use AI thoughtfully and with rigorous evaluation.
As Nick Brown mentioned in his blog last week, the U.S. Department of Education (DoE) recently published “Designing for Education with Artificial Intelligence: An Essential Guide for Developers,” aimed at guiding education technology teams - from designers to legal experts - in creating safe, secure, and trustworthy AI solutions in education. It was encouraging to see that the DoE’s guide aligns nicely with VitalSource’s own AI Principles.
When we began developing our AI Principles, I started by distilling the values we already held as a company and our mission to power the network that delivers transformative learning. We agree with an executive order on AI that “AI reflects the principles of the people who build it.” In the Learning Science team, we had already developed AI for automatic question generation (AQG) years prior. What we did not want to do is think of AI as a hammer and treat everything else as a nail; rather, we see AI as a tool to be used if it can help solve an educational challenge. For our team, our challenge was how do we bring formative practice to more students? We replicated Carnegie Mellon University’s doer effect research, and the learning science findings were clear: the benefits of formative practice should be brought to as many students as possible. AI was simply a tool that could help scale the generation of formative practice and bring it to millions of students.
With the need for AI to be based in learning science and research firmly established, our existing beliefs on the use of AI during the development of our automatic question generation (AGQ) system made clear several other principles. Accountability was expressly involved when we determined the type of AI we chose to use for our AQG system and how we maintained oversight. Transparency and explainability were cornerstones to our development process. We believed we should be able to explain exactly how our AI worked and committed to outlining the process in published research. Lastly, rigorous evaluation of the performance of the questions outputted from our AI system was critical, as ensuring the efficacy of these formative practice questions for students was of the utmost importance.
The next step to developing the AI principles was to understand the guidance from governments and key standards organizations. After reviewing materials from the White House, European Union, and organizations such as the National Institute of Standards and Technology, we were able to clearly distill what was most valuable to affirm for our educational technology context.
The final AI principles and definitions are an authentic reflection of our values here at VitalSource. Our development and application of AI – both now in the future - will adhere to these core principles, ensuring our continued trustworthy and responsible use of AI.