How qual researchers really feel about generative AI

The advent of generative artificial intelligence (AI) has triggered an existential crisis for qualitative researchers. Its mechanisation of knowledge threatens to undervalue our craft as it churns out increasingly credible discussion guides, hypotheses, summaries and even probes. Unsurprisingly, a 2023 paper from Princeton University said market research analysts are in the top 10% at risk from generative AI.  

In these changing times, we qual researchers must ensure that clients appreciate the value they get from our uniquely human talents of curiosity and empathy (both when using and not using AI tools). This means being less humble about how specialised our craft is.  

In the face of change, I got together with Insight Platforms and online qual platform Qualzy to lead a conversation on the topic along with 167 qual researchers from around the world. We explored expectations of AI’s perceived impact, its risks and opportunities. Here, I share some of what we discovered and how we can thrive in this emerging world. 

Overall, most ( 59%) of the qual researchers we spoke to feel optimistic, with  90% expecting a significant impact on analysis and 86% feeling positive towards summary automation tools. Only 21% expect a significant impact on project design and automated probing, while 56% feel negative towards automating probes (fearing a lack of nuance). Qual researchers want to benefit from AI’s efficiencies but also feel concerned it might replace us and damage our reputation if it underperforms.

Forgive my mixed metaphors, but generative AI feels like a double-edged sword of Damocles. It offers efficiency, but the more we use it, the better it gets and the more likely it will eventually replace us. It can feel like we are Uber drivers, supporting a technology that makes life easier, but ultimately wants to automate us. Are we co-conspiring to build a future of self-driving, humanoid researchers?   

I believe we need to embrace the efficiencies in AI in order to survive. And to thrive, we need to reinvest any time (and energy) saved and double down on the human skills that add most value to clients (but which AI struggles to deliver without us). That way, as AI improves, we improve with it, always staying one step ahead.

What can humans do better than AI? In his book I, Human, the psychologist Tomas Chamorro-Premuzic suggests curiosity, emotional intelligence and humility.  Let’s look at each one.

AI cannot be curious, which is essential for discovering new information. Great qual research is underpinned by curiosity; we hear something interesting and feel compelled to dig deeper until we discover a pivotal insight. If humans are ‘in the field’ asking questions, watching behaviour and actively listening, we have more opportunity to feel curious and explore the high potential comments that inspire fresh insights. AI doesn’t have this sixth sense. We need recognition for this skill, and we need to use AI tools in ways that feed off and enhance our powers of curiosity. 

AI can mimic empathy but lacks the humanity to understand emotion. It can’t process complex context, is limited to training data and misses non-verbal cues. Empathy allows us to build authentic connections with clients and participants, communicate with sensitivity, inspire emotional disclosure, encourage honesty, develop emotional models of understanding and tell compelling stories. Empathy means we spot emerging needs, which AI struggles to do, as it’s constrained by historical learning data. Again, qual researchers need to earn credit for empathy, and we must use AI tools in ways that are fueled by and enable empathy.

The problem is that we need credit for our uniquely human talents, but qual researchers are typically humble. The more modest we are, the less the nuances in our skillset are appreciated and the more AI tools seem like ample substitutes for expert consultancy.  We must do a better PR job for our uniquely human skills. 

First, this means ensuring clients appreciate the benefits they get from our curiosity, for example, in how we design questions, tasks, probes and AI prompts that uncover richer insights. Secondly, this requires showcasing how empathy underpins our design and interpretation, for example, by making our analysis more open so they see the depth of our working out. Thirdly, this means making directional recommendations; not just summarising, but actually interpreting. 

Qual researchers should use AI to be more efficient. But we must also ensure we get recognition for our uniquely human skills of empathy and curiosity. We must apply these unique skills when using AI to make sure we stay at the top of our game, one step ahead of what AI could do on its own or in the hands of amateurs. That way, specialist qual researchers will get stronger with AI and not just be replaced by it.  

 Tom Woodnutt is founder of Feeling Mutual.