Associate Director Sophie Henson reflects on the vital need for real, human empathy in research and how she likens generative AI taking on human insight to a sociopath using mimicry to fake empathy.
I just finished reading (for the third time – don’t judge me) Sociopath: A Memoir by Patric Gagne – a fascinating look at what it means to move through the world without the emotional scaffolding most of us take for granted.
What stuck with me wasn’t the sensationalism, but the mimicry – the ability to sound emotionally fluent without ever really feeling a thing. The performance of empathy – minus the substance…
It got me thinking about the world we work in - and the mad rise of generative AI tools that do something eerily similar.
They’re slick. They’re smart-sounding. And in an industry that’s all about more / faster / cheaper, it’s easy to get swept up in the speed and fluency. But here’s the thing: AI doesn’t know what it’s saying. Or why it matters.
It can mimic emotion, adapt tone, simulate empathy - but it can’t experience any of it. And without real feeling behind the insight, we risk making decisions that look 'right' on paper - and fall flat in real life.
Because underneath it all? It’s just prediction. Pattern-matching - not meaning-making. The shape of understanding, without the substance. Real insight needs more. It needs context, curiosity, care. It needs humans.
AI can help us go faster. But it can’t sit with ambiguity. It can’t feel the weight of a question. And it definitely doesn’t care.
That’s why we use AI to accelerate our working process - but we never let it lead. Because insight without empathy? Isn’t really insight at all.
If you let a sociopath run the research on what makes us human, it might sound convincing – but it won’t mean anything. Letting AI lead the work of understanding humans is like letting a sociopath write the rulebook on how we think and feel!
In the end, asking something emotionless to understand emotion isn’t just risky – it misses the point entirely.
