During a coffee break at the 2023 EduData Summit in San Fransisco, I was chatting to author Dr Jonathan Reichenthal. He’d just delivered his keynote, “Data culture eats data strategy for breakfast”, and we were discussing some of the points he had raised.

As it was May 2023, we were only six months into gen AI mania, and inevitably, our conversation turned to predictions of what the future of AI and its use in data might look like. Eventually, as we discussed AI’s use in data analysis for work, Dr Reichenthal mentioned the idea of a person setting up their own AI replicant that could respond to other people. Particularly through text and email, it might be possible to create an avatar with such fidelity that differentiation was next to impossible. Possibly, he joked, all email correspondence would be done by AI, with humans none-the-wiser.

While it was only brought up briefly, the idea of interacting with an AI without being aware of doing so seemed like some sort of violation. It turns out I’m not alone. Since ChatGPT’s explosion just over a year and a half ago, other AI tools have become popular, and those who it’s forced on aren’t always happy. I’ve come across countless stories of people booting AI assistants from meetings, or refusing to speak or interact unless it’s removed.

How exactly using an AI tool to annotate a meeting is different from recording it, I’m not exactly sure. The former elicits a lot more hostility than the latter, and I tend to agree. So does the EU, incidentally, which recently introduced its AI Act that specifies people should be aware when they are interacting with AI.

With AI now becoming a very credible way of recruiting students, the QSIM team wondered how this may impact prospective students’ perspectives of their interactions with a university. Selecting an institution has, for a long time, relied on human interaction and connection. Could the benefits of AI recruitment be outweighed by the downsides of reputation loss?