It has been a year since the release of ChatGPT, and educators are still scrambling to respond to this new kind of AI tool.
Much of the conversation has revolved around the double-edged nature of AI chatbots for educators. On the one hand teachers worry that students will suddenly cheat on homework with abandon, since chatbots can write essays in ways that are difficult to detect. On the other hand, though, educators see the potential of the tools to save them time on administrative tasks like writing lesson plans.
But in a recent working paper, a trio of education scholars say that these discussions are far too “parochial” and short-sighted. They argue that if the technologists building these new AI chatbots are right that the tools will quickly improve, then the technology will likely lead to massive shifts in knowledge work — including in academic research and the white-collar workforce — and therefore raise profound questions about the purpose of education.
“It just raises all these issues about what on earth are schools for?” says one of the paper’s authors, Dylan Wiliam, an emeritus professor of educational assessment at University College of London’s Institute of Education.
The paper imagines four possible scenarios for how generative AI, as the technology behind ChatGPT is called, might change society — and what those changes could mean for schools and colleges.
The goal behind the thought exercise is to get ahead of a rapidly changing technology, and to avoid what the scholars call the “worst-case scenarios” that could result. With that in mind, they close with a list of recommendations for how education and technology leaders can respond to try to best harness the benefits of the technology.
At times the paper is intentionally provocative. For instance, it imagines a scenario in which AI becomes so good at instantly creating learning tutorial videos and entertainment that people stop learning how to read.
“Literacy has been a relatively recent thing … and it’s actually really hard,” says Arran Hamilton, a director at the consulting firm Cognition Learning Group. “We have to co-opt a part of our brain that actually is generally used for facial recognition and we’re borrowing that to use for literacy.”
After all, the scholars note, some research shows that the recent rise of GPS technology and mapping apps on smartphones have led people to become less able to read maps without the tools. Could it be possible that within a few short decades reading may, as the paper imagines, “become as quaint as Latin and the Classics—things that we learn for bragging rights and the conferment of social status, but not in the least essential (or even useful) for day-to-day living”?
For this week’s EdSurge Podcast, we connected with Wiliam and Hamilton to talk through what this AI-infused world might look like, and how educators can start preparing. They argue that the recent executive order by the Biden administration on the safe development of AI is a good start, but that it will take more big-picture thinking to respond to this technology.
Listen to the episode on Apple Podcasts, Overcast, Spotify, Stitcher or wherever you listen to podcasts, or use the player on this page.