New tech, slow burn.

Why adopting innovative technologies like AI for teaching always takes longer than anticipated, and why being a late adopter can make for better digital outcomes.

If you’re anything like me, you’re quite glad the AI hype cycle roller-coaster has peaked a little and we can all take a little breather from, what felt like, a daily barrage of model and capability updates that started when Open AI dropped its first public version of Chat-GPT back in November of 2022.

Don’t get me wrong, the AI train is still speeding ahead furiously but I get the vibe that folks, in the education space anyway, are beginning to express fewer “Oohs and Aaahs” about what the next amazing thing AI can do. In fact many are trying to pump the brakes in order to slow down and just think. I think we’re all a little tired of being impressed (impression fatigue – is that a thing?)

Yes, LLM’s are even more powerful than just a few months (weeks?!) ago and yet more incredible magic-like prowess will be revealed soon, but how does this affect what university-based educators do on a daily basis? It’s becoming increasingly difficult to determine who to take direction from, how and where AI will evolve to next and what affect its actually going to have. Unfortunately, it seems that experts are only a few percentages better at predicting trends than those with limited knowledge. After all. the average expert is roughly as accurate as “a dart-throwing chimpanzee” at predicting events if you listen to Phillip Tetlock

And while we are at it, there are a plethora of other complex issues related to policy, process, enrollment and funding that are way bigger and more complex than what a few clever tricks from AI will gloss over.

Those not-insignificant issues aside, what can those teaching in higher education settings do on their own to try and surf the AI wave of innovation but at the same time not be entirely undone by it. And how do they build on the good digital practices they have developed over years of teaching in a blended or online environment?

Also, how do they accomplish this without also burning out through sheer frustration trying to rapidly shift to a new dynamic not of their own making? Sometimes, it does seem that individuals and organisations are too quick to espouse the benefits of AI in their programmes without any tangible evidence that it will work as intended.

In trying to address the fears of educators and navigate the new reality of AI in education it's useful to look back at how revolutionary technologies changed work and society in the short term. John Maynard Keynes even wrote about ‘technological unemployment’ as far back as 1930. Encouragingly, to date, these worst-case scenarios have not come to pass.

From the very first steam engine to the more recent advent of the internet, real and impactful change through application of those technologies on a broader scale only started having a meaningful impact many years after their initial introduction. Some took decades to make any meaningful difference to work and study, others up to 10 years.

So while there is most certainly evidence to support an accelerating uptake and adoption of technology generally (World Economic Forum - 2023), there is still a case to make for the slower adoption of realistic and impactful versions of that new technology. No one wants to be left behind and institutions of learning most certainly don’t want to be seen as slow to adapt, but there is something to be said for a more cautious and considered approach, especially when it comes to teaching.

Individual attention spans and limited cognitive effort are oversubscribed already so isn’t it in our best interest to use AI, presently at least, in mostly experimental and limited ways until we can confirm that its having a positive effect?

To clarify, this is not a call to Luddites the world over to storm and break the proverbial mechanical looms. AI will and is transforming our world, the way we work, live and communicate. But, as we have only too recently seen and continue to struggle with, social media, as just one example, has a very dark side that societies may never quite be able to solve now that it is ubiquitous. Genie, bottle etc.

Along with the many challenges in leveraging AI including hallucination, lack of context and bias? ?a recent New York Times piece (The Data That Powers A.I. Is Disappearing Fast) reports that AI may soon have a data providence problem that limits what relevant information and sources it can search. If learners aren’t acclimated to question the responses from generally available AI or think to verify information it provides there will be a far wider set of potentially ingrained negative behaviours to address.

So where do the solutions lie?

Firstly, in this writers opinion, based on years spent developing, facilitating and speaking on educator professional development, a late adopter approach to ed-tech is never a bad thing (If you’ve ever experienced the frustration of using the latest version of almost any LMS you will know exactly what I mean).

It takes an immense amount of effort and support to be the first or early at anything ed-tech so unless you have a large budget, time a plenty and a healthy dollop of organisational leadership patience be prepared to not meet every expectation. Indeed, if any at all. And considering you’re wagering student experience and performance on it, is it really worth the early adopter risk?

Likewise, for educators who feel that AI-based tools and ed-tech is being forced upon them, there is a lot of evidence to support a more considered approach to its implementation. That’s not a veiled dig at administrators and management but as the successful deployment of ed tech is primarily a team effort it’s important reminder that all voices need to be heard.

Too often we read online about this or that new ed tech initiative with a shiny new tool or some revolutionary app that produced no tangible results. The trash heap of promising digital learning technology rivals the actual Pacific garbage patch and, let’s be honest, everyone in education has an ed-tech misadventure story or three to share.

Secondly, as an individual educator, when was the last time you engaged deeply with a colleague, with a small group or as a faculty collective on ed-tech applications and how they have impacted teaching? Some organisations take this very seriously and review their approaches and tools yearly but it’s my experience that many educators just don’t have the luxury of time, don’t have it as a priority or are simply against the concept of change in their way of teaching.

I will resist the urge to wheel out one of the many tired tropes about how teachers could be replaced by others with more technical capability. However, it has to be acknowledged that, without a little commitment and reflection, the measured use and introduction of AI tools will be left to others to direct how it gets applied. That will not always to the benefit of learners or the discipline of those that push back against it.

Lastly, and related to the previous point, precious little collaboration and sharing of experiences of good ed-tech practices by educators is shared within schools, faculty’s or departments. Its often only as the result of external or irregular factors that successful and widely applicable digital teaching practices come to light. Given how big an impact AI is due to have, the sharing of experiments and experiences can help to craft a nuanced, impactful and longer lasting response in its application.

So, the onus is on us as active players in the ed-tech space to help shape valuable and positive experiences with AI for the benefit of learning in all aspects, but, at a pace and cadence that shows real progress, connectedness and encourages positive learner persistence. This requires a deeper level of reflection, validation and buy-in from all contributors to create a cohesive response.

Yes, it comes at the risk of being labelled a bit boring or slow but, at the end of the semester, who is going to thank you and respect you more? Your students or your social media feed?