Nearly fifty years ago, California futurist Roy Amara coined a widely quoted adage on the subject of technological change. It says: “We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run.”
This is a quote very much worth keeping in mind when it comes to thinking about how higher education should react to the growth of artificial intelligence and, in particular, how we grapple with educating students for an uncertain future.
One of the great challenges of the current age for colleges and universities is that there is a narrative out there that technological change – mainly, but not exclusively, AI – is rapidly altering the world of work and that educational institutions are not doing enough to prepare their graduates for this brave new world. I would argue that it is fair to say that universities and colleges have not rushed to make big changes to their curricula and pedagogy to accommodate or anticipate artificial intelligence – and also that this has probably cost institutions something in terms of public esteem, because parents in particular are pretty freaked out about the consequences of technological change with respect to their kids’ futures.
But on the other hand, it’s absolutely not clear – for the moment, anyway – that artificial intelligence is in fact changing very much with respect to the labour market. Sure, there is lots of evidence suggesting that business owners and senior management want to deploy AI at scale in the workforce (though clearly they are to some extent dissuaded by the fact that there are few companies that have figured out how to deploy it in a way that actually makes money).
Amazingly, there are very few surveys of employees which look specifically at how AI is being used in the workplace. There are lots of surveys about how employers wish that AI was being deployed, what they intend by deploying it, and whether or not they are seeing any value from it, but precious few that actually get to the issue of how the technology is being deployed and how it is changing the tasks that the labour force is deployed to accomplish. As a result, educational institutions have had very little information about the kinds of labour market experiences for which they should be preparing their graduates.
Until now, that is.
Last month, we at HESA Towers surveyed roughly 2100 university and college graduates 40 years of age or younger and asked them how they used artificial intelligence at work. We didn’t care about how they used it for personal tasks (getting restaurant or travel recommendations, for example); what we cared about was how it was being used to complete tasks at work and how – if at all – it is changing the nature of their tasks at work.
Broadly speaking, the survey has three big findings:
- Most graduates use AI to some degree, although a small minority consciously choose to avoid it. By far the most common uses are things which are fairly mundane – things like scheduling or drafting emails.
- The number who are using it for more detailed analytical or creative tasks is significant, but for the most part this is leading to a speeding-up of existing routine tasks. It is not eliminating any tasks or creating new ones, but it is giving employees more time to spend on creative/client-facing tasks. That said, there are a small number of intriguing uses that are non-obvious in nature.
- For the moment, workers are being left to fend for themselves with respect to AI. Only one in ten said they were actually being instructed by their employers to use AI to complete most of their tasks, while 70% said their employers gave them no instructions one way or the other about using AI for most tasks.
All of which suggests that we are still in the early stages of AI adoption and that, in fact, its impact on the world of work remains embryonic. We’re still at the “overestimation” stage of this technology.
For now. But like I said at the start, eventually we’ll get to the stage where we’ve underestimated it. That’s what we need to prepare for. The good news is that there is still time to get this right.
And that, of course, is why we produced this survey: it’s chock full of information about how graduates are using AI, what they think of it, where they want more training and where – on the odd occasion – the evidence suggests that there are some new, breakout uses of the technology. And we’ve got all of this broken down by field of study so that the insights are actually usable for the purposes of understanding how labour market trends can and should be incorporated into curricula – because the way health grads are using AI is different from students in business or engineering.
The data we have put together should be of great assistance to anyone who is:
- Currently engaged in curriculum reviews and wants to understand how the labour market is using AI skills.
- In career services and wants to provide advice to students about skills they can hone.
- Anyone responsible for setting learning outcomes.
The key takeaway here is that no institution is yet really “behind” on AI, for the simple reason that AI’s impact in the workplace is substantially less than the hyped would suggest. But, if past waves of technological change are anything to go by, change will accelerate. The time to prepare is now.
