This post is intended to capture a hypothesis I have about artificial intelligence in education. Much of this is speculation, but I’m not including a lot of hedge words and phrases for two reasons. First, simplicity. Second, I’m trying to be at least a little bit provocative here. I always appreciate feedback but that’s especially true in this case, because I want to get a sense from as many people as possible whether you think I’m directionally correct, way off, or somewhere in between. Your views will influence our thinking for the DLAC Ignite conference in Sacramento in February, future blog posts, and other programming. If you have thoughts on this topic, please send to john@deelac.com
AI is now recognized as a major issue in education, both in terms of opportunities, and challenges such as academic integrity, data privacy, and acceptable use policies.
This feels like a relatively recent development. Although many educators have accepted that AI is here, systemic efforts to help teachers adjust were slow to develop. But now we’re seeing significant activity among our DLAC members (Montana Digital Academy and Michigan Virtual are just two examples), as well as major efforts such as the teacher training being developed and delivered by Microsoft and the American Federation of Teachers.
But I still fear that most school and education systems are behind the curve on AI understanding and adoption, and this is especially important for our DLAC community for two related reasons. First, online/hybrid/blended schools are particularly vulnerable to AI impacts if they don’t plan and respond. Second and conversely, online/hybrid/blended are particularly well positioned to take advantage of AI if they do plan and respond.
There are two predominant themes for why I believe education is behind the curve, which I’ll call The Convergence (of Factors) and The Countdown (to AGI).
The Convergence of Factors
I’ve long been a skeptic of the potential of technology to “transform” education because I’ve seen many technologies overhyped, including augmented and virtual reality, gamification, and others. The “transformation” hype isn’t new; computers, TV, radio, and other technologies were previously pitched as the transformational technologies. Plenty of others have made similar observations.
I’m not nearly as skeptical this time, because of the convergence of three major trends: technology, parents, and politics.
- Technology: AI already appears to be potentially transformational in many ways that are starting to show. AI has certainly altered my daily work routines, conversations, and not just how I find information but how I think about and process key issues. In just the past couple of weeks there seems to have been some pushback on AI as some writers express disappointment in recent progress. But I think this is in fact a signal regarding how much we have come to expect of AI. I experienced this recently when I wanted to create a short video for personal reasons. I tried some text-to-video tools and was disappointed that the resulting videos took more effort, and weren’t as good, as I had expected. Then I noted to myself that I was disappointed that I couldn’t just type in text and have a really good video based on my idea in 30 seconds. Nobody would have expected this capability to exist a year ago! How did my expectations leapfrog reality so quickly?
- The second theme is what I’ll loosely call parents, by which I mean to encompass all the decisions that parents and families and students are making that are leading to enrollment declines in traditional districts, and enrollment increases in charter schools, private schools, homeschooling, and other choice options. (The enrollment increases/decreases are not 1:1 because declining school age populations are a factor in some areas.)
Parents are creating the demand for new options, and in some cases (e.g. microschools) also creating the supply. These new schools are far less encumbered by regulations and bureaucracy than traditional public schools. In over-simplified terms, with regard to regulations and bureaucracy, the order from least to most encumbered is homeschooling<private schools<charter schools<district schools. That is also the order in which organizations will have the flexibility to adopt and adapt to AI.
- Politics is the third piece that ties parents, options, and technology together, because depending on how you count, between about a dozen and two dozen states are on a path to allow for private educational entities to be paid via public funds at scale. This was always the missing piece with private schools and homeschooling; growth was slow or stagnant because there was a limited pool of parents who would pay out of their own pockets at levels similar to the revenue that public schools generate.
Observers of the private school vouchers and Educational Savings Accounts have noted that these new private entity funding options are all in conservative states. This is true, but incomplete, because some blue states have funding provisions that seem, upon close scrutiny, to have quite a few similarities to the private entity funding laws in Texas, Florida, Arizona, and elsewhere. The details require a separate blog post, but if the funding experiment in the red states is successful, there is a path for similar funding dynamics to spread to other areas of the country even if they don’t use the same labels as in the red states.
So, the convergence is (1) world-changing technology, plus (2) the proliferation of school options and choices driven by parents, plus (3) the policy funding private educational entities. This creates a flywheel effect in which the entities with more freedom to experiment currently are able to innovate, improve their offerings, “sell” those to parents, and attract public funding via parent choices.
But that’s not all. The second part is the Countdown to AGI.
The Countdown to AGI
Above I noted that schools and states have been responding to AI-related issues. The problem, however, is that most of the focus within education, including in our DLAC community, is on the current versions of AI. There is very little attention to what is coming sooner than many people realize: Artificial General Intelligence (AGI). AGI is a vague term, but one useful definition is an AI that can do anything that a human can do from behind a computer. It’s not far off—some version of AGI is likely to arrive within three years (i.e., 2028), and it may be sooner.
The implications of AGI for all of society and education will be big. For online schools, the implications may be profound. Much of what is being delivered asynchronously will potentially be able to be delivered by an AI that generates text and videos, personalizes the content, tutors each student, and can generate and score endless assessments on the fly. Personalization that has been dreamed of for decades is likely to actually materialize. The role of teachers—already so different in online schools compared to traditional schools—could shift again.
Some educators, seeing this future, are arguing that the importance of relationships in teaching will increase, and that AI can’t replace those relationships. On the first point, I completely agree. On the second point, I worry that what educators and parents hope for—that students will prefer human interaction to conversations with AI bots—is already being called into question by the growth of options like Character.AI.
Looking to 2028
If 2028 feels like a long time into the future, ask yourself if 2022 seems so long in the past.
That’s the point at which interacting with an AI tutor via audio and text will be indistinguishable from interacting with a human tutor.
That’s the point at which an AI will be able to instantly create the content for an entire online course, personalized for every student in the class.
That’s the point at which it will be obvious that every school will need to prepare students for a world that their teachers and school leaders have never experienced.
I may be off in using 2028 as the target year. But unlike other technologies such as driverless cars, AI is developing faster than most predictions.
What happens in this world? The schools that have the most flexibility and foresight will flourish, provided they have the funding. Schools with fewer legacy technology systems, employment contracts, state requirements, etc. will be advantaged, because every one of those elements is a drag on moving quickly, and in a world of funding following educational choices, the schools moving more quickly will be rewarded.
What might these innovative schools do? Here, we need to separate primary from secondary schools, because the approaches are likely to be very different.
At the elementary school level, the (self-proclaimed) AI-powered Alpha schools are already demonstrating a model. Students can learn from AI content and tutors much more efficiently than they can from teachers—if teachers and other professionals are providing guidance and support. The efficiency of instructional time will allow the school to focus on topics and activities that fire students’ interests and curiosity. Cost structures will shift, with paying for teachers becoming an even larger part of school budgets—if state legislatures don’t respond to falling technology costs by cutting education funding. Schools will be better able to accommodate parents who seek different levels of involvement in their children’s education.
At the high school level, successful schools will build on students’ interest and agency to provide them opportunities for pursuits that will include jobs, internships, college courses, and other interests—much as innovative schools are already doing, but at a completely different scale. But this will only work if all schools receive funding for supporting their students in these new endeavors.
At both the elementary and high school level, schools that adapt in this way will look quite different than they do now. This is where the greatest threat lies. If we have parallel educational systems in which public schools have to compete with publicly funded private entities, while being encumbered by legacy regulations, we risk a digital divide that will make the digital divide of computers and Internet access look like an inconsequential trial run.
Because this is a system in which the parents who care about education the most, and have the time to pay attention, seek out the innovative schools. The students who don’t have that level of family support will be enrolled in the schools that remain mired in regulations and bureaucracy.
Let’s discuss
Are these ideas on your agenda? Is your school leadership and school board thinking about these ideas? Your regional service center and state agencies? Should any of these people and entities be putting time into these issues?
Alternatively, do you believe it’s not going to happen soon enough to worry about—or maybe these changes aren’t going to happen at all?
Finally, if you think this is important and there’s a decent chance that the future that I’m proposing may happen, and you don’t see the preparation and planning occurring, what should DLAC do?
I would love to hear from you. john@deelac.com
(This post is planned to go out the day before I go mostly off-grid for a week. You’ll get my OOO, and you won’t hear more for a while, but I will be in touch with everyone who emails.)
We welcome your comments in the DLAC Community Portal’s Blog Discussion Group. Scroll down to the bottom of the page to join the conversation.