AI in education: sooner and later
It’s hard to ignore all the attention to artificial intelligence generally, and how much AI in education is being discussed.
In fact, there’s so much out there that the problem isn’t a lack of information, it’s understanding what’s real, what’s hype, what you can and should do now, and what you should have in mind for later. This post breaks down some key ideas and sprinkles some suggestions throughout.
Current status of generative AI
AI has been in development for decades, but with the release of ChatGPT just over a year ago, generative AI emerged and became widely and publicly available for the first time. Since then, ChatGPT became the fastest-adopted technology in history, while other generative AI platforms have become widely used as well.
Why has adoption been so quick? Two main reasons seem apparent. First, the capabilities of AI platforms are astounding. Already, ChatGPT performs very well on various high school and college tests, with the paid version scoring at or above the 80th percentile of test takers on various AP exams, the LSAT, GRE, and other assessments.
That of course is just one measure, but it’s related to the second reason for such rapid adoption: it is very easily used. It takes minimal effort to use ChatGPT to write a first draft of essays, lessons plans, job descriptions, etc. The output isn’t perfect, and has to be checked and refined, but nearly everybody who tries it finds that it saves them considerable time on a variety of tasks.
Our first suggestion: if you’ve not already started using ChatGPT or other AI platforms, do so even for just one task per week. There is a large and growing divide between people who are using AI in any consistent way compared to those who aren’t.
Current concerns about AI in schools
For every article exploring the promise of AI In education, there seems to be at least one article concerned about improper use of AI by students, related to academic integrity or other reasons.
These are legitimate concerns, but they are also limiting.
Many technologies, from calculators to Internet search engines, have created concerns about academic integrity. That doesn’t make such concerns less important, but it does place them in context. Such concerns have risen before, and they will again, and solutions are available that mix technologies to combat cheating with the better approach of increasing the understanding of teachers of each student’s academic status and progress. A few states such as Oregon have released guidance on AI for their schools, and many organizations are suggesting responsible use policies as well.
In sum, these are important interim issues that teachers and school leaders should keep in mind. But if that’s the extent of your focus, you will remain perpetually in catch-up mode for the reasons explained next
Our second suggestion: Treat understanding and addressing student cheating and acceptable use as the floor of your activity related to AI, the starting point, not the ceiling. And don’t get too bogged down in it and distracted from the bigger issues coming.
Current and future capabilities of AI in schools
Perhaps the main use of AI in schools currently is by students for a variety of legitimate and dishonest uses. But that is changing in ways that are accelerating. Already we see significant progress in several areas, including:- Lesson planning: An Internet search for “AI in lesson planning” returns page after page of guidance, platforms, companies, and other avenues to assist teachers. Many are free, others are already trying to sell a product. Of course, as with any guidance or product, some are better than others. It’s hard to know without a lot of effort which are the best sources—but it’s clear from the guidance and examples, that AI is being used to plan lessons now. It’s also clear that some teachers understand the capabilities and limitations of AI, while other teachers risk an over-reliance in AI that could lead to bias, inequity, and blatant errors. As suggested above, AI is currently best seen as a first draft, not a finished product.
- Content development: It’s not just lesson planning; it’s also the content of lessons themselves that increasingly can be developed by AI. This is less common among teachers and in school districts, but it’s already happening in content development companies.
Content creation is in fact an area in which AI is likely to shine. Companies already have to create different content for different state standards. Teachers already adjust to the needs of individual students—although the time they have to do so is limited.
Not far in the future, we can expect AI platforms that can instantly transform a curriculum to better meet the needs of, let’s say, and English Language Learner. Or a student with dyslexia. Or a student who is reading ahead of their grade level. The possibilities are nearly endless. - Tutoring: Educators understand the power of a one-on-one discussion between a teacher and student. That’s why many districts invested stimulus dollars in tutoring, and why one-on-one instruction is considered a high impact intervention.
AI isn’t yet providing tutoring at scale, but emerging signs exist, with Khanmigo, private schools relying on AI for instruction, and the interest in AI “characters” that students interact with.
Our suggestion: when you are considering content or technology providers, asking “what are your current AI capabilities and future AI plans” should always be part of your vetting process.
StrongMind can help
Does this all seem daunting?
If your answer is “no, I understand all this,” can we book a meeting with you to explain it all to us?
Because the honest answer about AI in education — and society more broadly — is that nobody knows where AI is going to take us. But at Strongmind, we’re on a path to support the growth of AI in education while keeping the focus where it should be — on teachers and students.
Please contact us or join our watch a recording of our webinar if you’d like to learn more.