AI edtech experts? Show me the evidence

Dan Sandhu, CEO of Sparx, discusses his concerns around the lack of bona fide AI experts within edtech

In June the Department for Education (DfE) announced the launch of its AI-horizon scanning group to explore the promise of artificial intelligence (AI) in education. The remit of this group isn’t immediately obvious but is loosely described as bringing together “digital, policy, and delivery professionals to explore how artificial intelligence might impact the UK’s education system”. As an edtech professional very much engaged in exploring AI in personalised learning technologies, perhaps I should welcome such a group wholeheartedly. And, while any such focus on investigating how to improve learning is positive, my worry is that such expert groups sometimes miss the most important facet of their remit – interrogating empirical evidence. 

You might also like: UK government announces £18.5m investment in AI training

Anyone who takes an interest in edtech will no doubt have read many a comment piece either extolling the virtues or questioning the impact of AI in education. These opinions are passionately and persuasively argued, providing personal experience and insight into the issue, (not least from suppliers of AI-empowered edtech solutions such as myself). However, it seems that there are two important elements missing from the debate as it stands.

The first concerns edtech evangelists’ seeming willingness to put all their eggs in the AI basket. In much of the commentary I see, AI is positioned as a panacea to any number of classroom challenges. Similarly innovative, alternative technologies seem to be sidelined in such commentary in favour of AI’s buzzword-based pulling power. The second missing link in this debate is the impersonal – that is to say, empirical, unbiased evidence based on rigorous research rather than personal passion.

It’s that lack of evidence that is most concerning. As Nick Gibb, schools minister, said in a response to a written parliamentary question: “AI is a complex, emerging area… However, the impact of these technologies in the classroom still remains largely unevidenced.”

 We only hope… that the current political turmoil does not take focus away from education’s need for innovation based on evidence.

The Nuffield Foundation’s recent report, Growing up digital: What do we really need to know about educating the digital generation?, echoes these concerns: “Any research programme will need to take account of the need to gather data from established practice, not simply at the implementation stage where technical glitches and the halo effect can both skew findings significantly.”

Research is time-consuming with few shortcuts. It requires edtech suppliers to invest in their own rigorous research and evaluation, which can be costly. However, the DfE’s recent support for a more evidence-based approach to edtech suggests there is greater political will to invest in research programmes. Small studies may well be able to inform this, but without a large-scale effort the UK will fall short in its understanding of whether edtech can really make the impact we hope.

When it comes to AI a similarly sober, long-term approach needs to be championed – not to quell the desire to innovate, but to ensure that AI implementations couple passion with evidence. It needs to make a measurable difference to learners and schools.

You might also like: Buckingham VC says AI is a ‘bigger issue than global warming’

Recent research from MMC Ventures suggests that 40% of ‘AI startups’ don’t actually use AI. Now consider each of those startups are led by founders positioning themselves as AI experts. In the excitement surrounding AI, it seems that the bar to qualify as an expert has lowered considerably. I’ve been involved in setting up and running edtech startups for over 10 years and it’s clear to me that we’re in something of an AI bubble at the moment. AI is almost like catnip for investors and Google searches for the term have increased significantly. In my experience, this leads to overblown claims and a reduction in expert research-informed debate. 

As the DfE gathers its AI steering group, I hope that robust evidence and a programme of research is its number one priority. We only hope that the path set by this administration’s strategy continues under new government leadership, and that the current political turmoil does not take focus away from education’s need for innovation based on evidence.