Institutional vs consumer EdTech: Where will AI have the biggest impact on study?
Photo by Avelino Calvar Martinez from Burst
"Online learning is fundamentally broken” is a phrase I’ve encountered multiple times since the release of ChatGPT in late 2022. While questioning the efficacy of online learning is as old as online learning itself, it’s been interesting to observe how the AI explosion has led some into discourse about the state of online learning.
Online education isn’t the only thing that has experienced some “crisis framing” since that time; education itself has been framed as something that needs saving.
Unsurprisingly, one reason for the uptick in crisis framing is that people want to tell you how AI can fix online learning or save education. Essentially, there’s a lot of AI-led problem-solution marketing going on.
Some of this relates to products, and some to people who, I have no doubt, have used the approximately 13,000 hours that have elapsed since ChatGPT was released to reach the Gladwell threshold of expertise.
We’re at a stage where people want to gain some kind of market share, whether for new product lines or for attention and professional prestige. One way to create fertile ground for that is to simplistically problematise higher education or online learning, or to heighten the sense of the need for action… think of the “AI has, or is, going to revolutionise education” type arguments.
This sometimes feels more like a fear-driven approach than a positive one and taps into a spirit around technological developments that Jacques Ellul seemed to hit on in his book The Technological Bluff when he said:
“We live under threat that if we fail to embrace new technologies, we will be pushed aside into cultural obsolescence, left without key skills to get a job, disconnected from cultural conversations”
There’s an interesting social psychology aspect to the maelstrom of discourse around AI and its impact on education. That maelstrom can cloud perceptions of both present and plausible near-future reality.
Can we really say that AI is revolutionising higher education? That is to say, it is changing higher education radically and fundamentally. I don’t think we can say that right now. If we look across a range of dimensions, there is not enough evidence to suggest that UK higher education has radically changed since November 2022.
AI may well be the catalyst and author of a total and demonstrable transformation of higher education, but that transformation is yet to occur.
AI developments' impact on higher education and online learning is, however, an important topic. But it’s one in which there’s tension between what might be possible, what is probable, and by when.
We may be able to think up a whole array of compelling ways in which AI can be used and implemented, but there’s potentially a big gulf between those ideas and higher education institutions (HEIs) and their staff being sold on them, being willing and able to implement them, and/or there being the wider conditions in the sector to bring them to fruition.
Of course, AI isn’t like interactive whiteboards, for example. It’s not exclusively a specific thing you deliberately procure; it might be that in terms of AI-heavy products, but it also has and will continue to permeate digital technology already in use at HEIs. So there’s much greater subtlety here because where there’s digital technology, there is and will be AI.
However, there are differences in how AI “gets in” to higher education, if I can put it as crudely as that. So I want to consider two contrasting ways that AI features and technology will influence higher education study.
AI from the academy
The first area is new AI-driven product features and developments being implemented within the education technologies already deployed in HEIs. In a UK context, these include virtual learning environments (VLEs), video conferencing technologies, digital assessment tools (or Turnitin as it is also known here), and several supplementary technologies that form part of a relatively uniform core collection of edtech products.
Product developments within existing technologies used by HEIs will be one of the main ways in which AI will continue to impact UK higher education. The main conduit is likely to be through new features and functionality within the four main VLE products: Blackboard, Brightspace, Canvas, or Moodle.
All the companies behind these products have announced, implemented, or are developing AI-based features. However, what's interesting is that they are largely targeted at educators, not students. All four VLE products have announced features that include using AI to help structure or template courses, or to help create assessments, quizzes, and rubrics, but there has been a lot less focus on features that students use.
Brightspace, with their virtual assistant that offers students contextualised help and documentation from within the product, is one student-focused example. Another main one is Canvas integrating Khan Academy’s Khanmigo, an AI tutor initially focused on helping student writing by acting as a writing coach. But apart from those examples, there have been few directly student-focused features.
This brings us back to the gap between possibilities and realities. Currently, UK higher education as a whole is best described as being cautious about AI, with a tendency to focus more heavily on perceived threats, ethical issues, and negative/unintended consequences. That inevitably influences product development, which in turn affects what’s available in terms of AI features and functionality.
There have also been examples of direct influencing of products, such as the lobbying in UK HE to turn off the AI writing detection features within Turnitin. Although, let’s be kind and just say that this is an area of particularly low efficacy.
So while existing technologies are one obvious way in which AI features will permeate UK HE study, the sector’s disposition is undoubtedly influencing product developments and, in some cases, there is pushback on product developments.
Although there are and will continue to be more AI-specific products developed that HEIs could procure and start to use, there has yet to be a new, explicitly AI-centred EdTech product that has been widely adopted by UK HEIs.
The direct route
AI technology available now and in the future to HE students is an area of real interest and brings us into interesting territory. Technologies procured and supplied by HEIs are complicated by the need to serve two main audiences: educators and students, as well as other types of HEI staff. Their development is also influenced by the factors I’ve just mentioned.
However, direct-to-consumer technologies do not face the challenge of catering to different types of audiences. Essentially, it's the direct-to-consumer AI technologies or the embedding of AI features into more generalist technologies likely to be used by students that have caused higher education so much anguish since ChatGPT arrived, most specifically regarding how ChatGPT might be used to generate essays.
Direct-to-consumer or student edtech is an interesting area and one that has seen the rise of companies like Chegg and Course Hero in recent years. Whether directly education-focused or not, AI direct-to-consumer products are where the main impacts might be seen over time.
The term Personal Learning Environments (PLE), which came to prominence in the 2000s, feels very old school now. This was essentially the idea that students might develop a collection of technologies that form their own digital learning environment. However, AI technologies now offer students an array of more sophisticated tools to support their learning and study, and it’s clear that many students are using those technologies.
A HEPI survey of 1,200 undergraduate students published in February 2024 highlighted that 53% of students had used generative AI to help with their assignments and 36% used it as an AI private tutor. We’re also seeing new acronyms such as BYOAI (bring your own AI) being used, that highlight the personal adoption of this tech.
There have always been technologies available to students that fall outside the institutionally supported collection, but things feel different now in respect to AI. There are, of course, moves to sell major products into universities like ChatGPT Edu and partnerships being developed such as Grammarly’s partnership with the large US online university Western Governors University (WGU). But the HEIs in the news for these kinds of things are the usual suspects like Arizona State University (ASU), and they form a mere fraction of all HEIs. This skews the reality of the wider picture somewhat.
I may be proved wrong on this, but I think the direct-to-consumer product space, whether deliberately education-focused products or not, is going to be the area of most interest in terms of how AI influences higher education study. It may also become the area that edtech focuses on more heavily, given how difficult it can be to sell new products into HEIs. It’s also going to be an area where we might continue to see growing disconnects and an anachronous juxtaposition of HEIs and their students.
A two-track blend
Those who scan higher education news regarding AI developments risk amplifying their own availability bias, creating a false understanding of what is actually happening across the whole of higher education and what will happen in the near future.
AI, of course, offers new possibilities for educators and students in HEIs and creates ripples within the education landscape as a whole. But the extent and speed at which things will actually change are unknown. However, what we do know is that there’s complexity across the broad HEI landscape and a spectrum of positions regarding AI, ranging from outright rejection to unbridled advocacy.
The area with the fewest barriers for some students (particularly those with financial means) and arguably the most fascinating area is how individual students will start and continue to use AI technologies alongside their study and teaching.
This is where we might see the most fundamental changes over time and ultimately continuing challenges for HEIs. While AI developments are being seen within the educational technology currently in use at HEIs, this progress is slow and cautious, reflecting the sector's current aggregate feelings on the subject.
Ultimately, the impact of AI in higher education will likely emerge through a blend of cautious institutional adoption and student-driven use, with the latter being an area worth watching more closely than it ever has been in the past.
