Artificial intelligence in education: reality and hype
Artificial intelligence (AI) holds promise to improve education, but as yet is over-hyped and misunderstood.
Let’s start with the over-hyped. Examples abound, but I’ll just point to this one: 26 ways that artificial intelligence is transforming education for the better. A brief look at the headline would lead one to think AI has taken over education, but the article has few good examples and fewer links or citations. It is mostly made up of assertions such as
“Classroom/Behavior Management. AI is currently being used to help teachers manage student behavior and the entire classroom.”
That is the entire assertion, with no documentation or citation. Although a high school English teacher would reject writing at this level, these are the types of articles that grab attention and lead people who aren’t close to education but are influential, such as state policymakers, to mis-understand what is truly occurring.
None of the examples in that article show AI being used extensively in education. Some of the examples are promising, but not yet being implemented at scale. Other articles seem to conflate any use of computers or technology with artificial intelligence, further confusing the picture.
AI is misunderstood, in part, because it can be used alongside terms including “big data” and “adaptive learning.” But big data and adaptive learning are both vague terms that are often used to describe actions that don’t require AI, and most current applications of adaptive learning and data in education do not use AI.
In this mass of confusion, a January report from Rand is incredibly useful. Artificial Intelligence Applications to Support K–12 Teachers and Teaching: A Review of Promising Applications, Challenges, and Risks, presents a sober, reality-based assessment of the current state of AI in education. The key takeaways:
“the influence of AI applications in the education sphere…has been limited… Although various AI advocates are currently touting a myriad of new applications for K–12 education, there is little evidence yet to support the usefulness of these applications to districts, schools, and teachers.”
The report then identifies
“three areas in which AI-based solutions have shown promise for supporting teachers in challenging areas of instruction: adaptive instructional systems that allow teachers to differentiate instruction at the student level for certain topic areas and skills; automated scoring of student writing assignments, which supports teachers’ ability to assign more writing in the classroom; and early warning systems, which alert administrators and teachers when students may need additional support to stay on track and progress toward graduation.”
The author doesn’t suggest that these are areas that are using AI widely, but that they are the most promising, and concludes that
“The most-effective AI applications will continue to play an assistive role, supporting rather than replacing teachers in their work with students in a limited set of content and topic areas that are most amenable to AI approaches. The most prevalent use cases will continue to be blended forms of instruction, in which the use of AI applications is integrated into teacher-led instruction and classroom activities.”
The study also includes useful definitions and explanations of AI, creating a very meaningful primer for anyone trying to understand the basics. In these and other ways, it provides the understanding needed to interpret poorly-researched articles and counter the AI hype.