A perspective from Fulcrum Research Group
A recent article from bi.team explored an idea that immediately resonated with us: people tend to believe AI will transform other people’s jobs more than their own.
Their research, based on a survey of more than 3,500 working adults in the UK, found that 43% of respondents believed AI could perform most of the tasks involved in other people’s jobs within the next ten years, while only 35% believed the same about their own. Across every demographic they studied, a similar pattern emerged.
We think this is an important observation, not because it tells us who is right or wrong about AI, but because it highlights a challenge that every organization introducing AI is likely to face.
At Fulcrum Research Group, our conclusion is slightly different from the article’s. While involving employees in AI implementation is essential, we believe there is another practical way to bridge this perception gap:
Give people the opportunity to weave AI into their workflow to improve it before thinking about whether it offers a wholesale replacement.
Most conversations about AI focus on capability. Those are important questions, but they’re rarely the questions people inside organizations are actually asking. Instead, they’re wondering whether AI can understand the nuances of their work, whether they can rely on its outputs, and where it genuinely adds value.
These questions don’t have theoretical answers. They have practical ones and practical answers come from experimentation.
We all see work differently
One of the strongest points in the article is the distinction between looking at a job from the outside versus experiencing it from the inside. Outsiders see tasks; insiders see judgement, tacit knowledge, relationships, and context.
This naturally leads people to think AI cannot replicate what they do. Sometimes they’re right. Sometimes they’re surprised. The only reliable way to find out is to test.
Rather than asking ‘Should we use AI?’, a better question is ‘Where can AI genuinely help us?’
A Phase 0 for AI
At Fulcrum Research Group, we’ve been exploring exactly this challenge through Synthetic Users – AI-generated physician and patient personas designed specifically for healthcare market research.
Rather than replacing primary research, Synthetic Users provide a rapid way to explore ideas, challenge assumptions, refine messaging, pressure-test concepts, and generate hypotheses before engaging real respondents.
We think of this as a Phase 0: an opportunity to experiment with AI in a safe, evidence-based way before making larger decisions (such as which concept to test in global research programs).
The goal isn’t replacement
History shows that technology changes jobs instead of eliminating them. We believe market research will follow a similar path.
As AI takes on repetitive or exploratory work, researchers have more time to focus on asking better questions, interpreting emotion, understanding context, and providing strategic advice.
These human capabilities remain central to high-quality research.
From perception to proof
The bi.team work concludes that successful AI adoption depends on bridging the gap between leadership’s perspective and employees’ lived experience. We agree.
We suggest there is a straightforward solution: experimentation.
When teams can use AI alongside their existing workflows, the discussion shifts from speculation to informed decision-making.
That’s the role we see for Synthetic Users — not as a replacement for researchers or respondents — but as a complementary research method that helps organizations explore ideas earlier, learn faster, and enter primary research with stronger hypotheses.
Closing the AI perception gap isn’t a binary choice between can it do the job or can’t it.
There is greater opportunity to be had by weaving it into our workflows as tools and roles evolve.
