“We have to reinvent ourselves to stay relevant,” my colleague Francis Hintermann told a room full of research folk in London last week. It was the defining theme of the 2026 KIMRA Conference – and reminded me (with a timely prompt from Stephanie Cook) of the mice in Who Moved My Cheese? by Dr. Spencer Johnson. In short: you can waste away grumbling about change (like someone moving your cheese) or lean into it (find out where they’ve put it).
I think, therefore I am employed
On my panel, which gamely tried to divine the future of the research function, I suggested the audience ask themselves and their teams a simple question: are you employed to deliver or to think? Sure, they are not mutually exclusive – but which has primacy?
If you believe it’s delivery then you’re competing against generative AI tools like AlphaSense that can whip up well-structured, deep research reports in minutes. That’s a battle you’re going to lose.
But deep, creative thinking remains human. As such, we researchers should apply our grey matter in three specific ways to thrive in a world where ‘anyone can do research’:
Skepticism
The scenario is probably familiar: a colleague – who seems to have just discovered generative AI tools – sends you a long document of grammatically tight, structurally sound prose. “Could you just check the numbers and sources are OK?” they ask.
The first advice I’d give is to take a couple of deep breaths. The second is to adopt the role of a skeptical advisor; become the arbiter of discernment and taste. To put it bluntly: call out the bullshit.
Point out what is accurate and what is fabricated. Show how good points may be hidden beneath generalisation. Ask: “What does this tell us that we did not know before?” Explain how some deeper thinking could be overlaid to get you to a more insightful place.
Knowledge
A key danger presented by generative AI tools is “cognitive offloading” – and a corresponding collapse in critical thinking. An MIT study found 83% of students who used LLMs to write essays unable to subsequently quote from them. If we don’t engage, we don’t remember.
We researchers must not be lured into this trap. If we sacrifice deep thinking on the altar of speed then what value are we really adding? It is our knowledge banks that will set us apart; the ability to weigh up conflicting points of view; to assimilate and then synthesis disparate information.
To that end, we must continue to read (long-form) and also to write. As my former colleague Dave Light eloquently puts it: “writing is thinking…a process of discovery – a process of finding out…what seems brilliant one day may look weak the next.”
Strategy
Every researcher has been challenged on pace: to speed up design, to speed up fieldwork, to speed up analysis and reporting. From synthetic respondents to analysis ‘at a click’, AI tools promise to do just that – and more.
However, speed gained to the detriment of strategy is a pyrrhic victory. With all these capabilities at our fingertips, it’s all too easy to plunge in without properly exploring the most critical element of any research project: the question we’re trying to answer.
Getting that critical initial question right is often the hardest part. Don’t let the obsequious chatbot fool you into thinking you’ve landed on something novel on your first pass. Speak with colleagues; hunt the whitespace; iterate, iterate, iterate.
More fox, less hedgehog
Fellow panelist Andrew Cosgrove offered a simple, compelling message at KiMRA: “Be less hedgehog”. These prickly mammals have failed to adapt to their environments; rolling up into a ball deters a badger, but not a car. Similarly, the researcher who sits still risks being flattened by AI.
Hedgehogs were famously contrasted with foxes by Isaiah Berlin – more adaptable animals that have evolved as the world changes around them. The smart researcher will do likewise: seek out diverse opinions, integrate new data, champion unpredictable outcomes.
AI must be a tool to take insights to new depths, not a crutch to do the ‘same old’ a little quicker.
