LLM EdTech Will Accelerate Progress in Education

Why has progress in education been generally slow—at least as measured by standardized international student assessments—compared to the rapid advances in technology over the past two decades? Why have learning management systems (LMSs), ubiquitous access to mobile devices and the availability of vast knowledge resources on the web contributed little to close the education gap between students from both higher- and lower-income backgrounds?

While there are many factors to consider, the single most effective way to boost the rate of learning progress is one-on-one tutoring, which, of course, has not been affordable for students who would benefit from it most. Large language model technology (LLM), with its ability to converse like a human, promises to close this gap, providing a personalized (yet fully automated) AI tutoring experience for every child, in all common human languages.

Although standalone LLMs can be easily prompted to chat like a middle-school tutor, additional technology components are needed to achieve the full potential of this vision: an integration with Wolfram tools for AI to get the math and facts right (LLMs are weak in computation and tend to invent knowledge), long-term memory between sessions, generation of practice examples, automated assessment methods and data-driven personalization engines. LLMs will need to be tuned for this purpose and optimized for cost-efficient cloud operations. An essential part of this platform will be the ability to host custom, localized curricula, as well as assessment standards. Last but not least, data privacy and security need to be part of a sound plan.

To stay ahead of their competition, EdTech companies worldwide need to rethink their offerings and business models in order to develop a future-proof strategy for a world in which everyone has access to personal artifical intelligence. Navigating the rapidly evolving AI landscape and making business-critical investment decisions is a considerable challenge for anyone in education technology. While Wolfram’s R&D teams are already building the technical foundations of future AI education systems, Wolfram Consulting Group can help you now with developing an overall strategy, creating attractive product concepts and rapidly building prototypes of your future products and services.