Wherever you sit in this sector, if you don’t have a perspective on AI that you’re constantly ruminating over, advancing and refining, I envy what I assume must be the confidence born of being an early mover and a singular authority.

Most of the rest of us are applying a portion of our brains – hopefully not the fear-sensing amygdala but the regions associated with creative thinking – to figuring out where this profound disruption takes us, where we go next and what it means for us personally and for our businesses.

It’s a work-in-progress exercise that requires test-and-learn loops, collaboration, open-mindedness and critical thinking. It presents a challenge many of us have not experienced before. While some have drawn analogies to previous disruptions, such as online surveys and focus groups, these don’t come close to how AI is starting to transform our work.

For our brains, there’s a peculiar challenge in all of this, whereby we have to operate on several planes at once – figure out a stance but be prepared to adapt it, given this is a fast-moving space; hold on to quality standards without being defensive; engage expansively and imaginatively while keeping a critical distance. It’s exciting, but we’re also allowed to find it all rather tiring sometimes.

It’s relatively easy to be a rejecter, and while it’s not easy, your position is clear if you’re an innovative first mover. The vast majority of us are at neither extreme, but figuring out real-time how we thrive in this changing landscape, in the messy middle.

There are many positions – intellectually and practically – on the spectrum between the rejection and enthusiast extremes. I’m a researcher, so for all that I’m loving the possibilities of AI chatbot moderation, I often learn in dialogue. So, I wanted to talk to people across our sector to understand their perspective on what AI means to them, and what they think it means for research and insight.

Brilliantly, there’s complexity and nuance in every person’s perspective, even those who are most enthusiastically leaning into the revolution. All views below are their own, rather than a company stance. Several themes emerged.

The industry will be carved up in new ways

It’s not hard to be cynical about synthetic data because there are various case studies out there showing how it doesn’t deliver versus real human beings – but of course, ‘traditional’ surveys can have their issues too, and while many hold a cynical view on this space, it is evolving. Leanne Tomasevic at Electric Twin talks about the confidence she has in the company’s approach, while recognising that quality requires good seed data. On synthetic data, she notes: “It’s playing a role in the gaps that traditional methods don't currently fill, and why wouldn't you want that?”

Clients are, of course, already engaging with synthetic data or are likely to be actively looking to do so. Fenny Leautier at Philips is actively exploring synthetic data, while being cautious to uphold quality standards and ensure care over issues like data security. In the wider context, the use of synthetic respondents is currently a polarising issue across the sector, and as in-housing continues apace, there’s potential for the providers of bespoke ad-hoc work to be cut out of some opportunities, and for tech players to edge into new value spaces. This can create new types of opportunity client-side. Leautier says: “We had an open position recently, and we made it a consumer insights AI platform manager, then we changed one of our leader’s roles to focus on this area, too. To make this work properly, you do need insight experience to evaluate if quality standards are being met.”

Jake Steadman at Canva notes that the sector is overall only superficially integrating AI in the main today, and feels the excitement in new entrants innovating in AI-led propositions. He wonders if, as researchers, we are holding too much on to our commitment to robustness, rigour and the significance of our work, and need to radically rethink what value we deliver, and how – with less focus on methodological purity and more on speed and agility.  

Steadman comments: “A lot of researchers rightly care about craft and getting to the truth. They ask good questions about robustness, which is obviously super important. But sometimes that pursuit of certainty slows us down. When we have a strong signal and confidence in the work, we should trust it and move faster, without losing sight of the truth.”

Meanwhile, Russ Wilson at The Sound reflects on this year’s MRS annual conference, noting that it felt like there was more emphasis on tech providers than ever, and less on the research itself. He questions what the sector really looks like now – and if we all have a common view of this – and what it’s going to look like in the future. It’s certainly reshaping in ways we might not quite know how to reconcile.

How do we conceptualise AI?

Anneke Quinn-de Jong is a former client-side insight professional who works as an AI transformation consultant, currently with Schiphol Airport. She says: “A lot of people still see AI as a tool, and if you see it that way, it’s one more thing to learn. But in my opinion, it’s going to have a much bigger impact than just a tool that we use. It’s not like a new version of a Microsoft application. It’s going to be much more disruptive than past work revolutions.”

She feels we’re only seeing the tip of the iceberg and should be prepared for much more change ahead. She recognises the risks we are facing while we embrace opportunities and believes it’s better to control our own destinies by taking ownership than push back and have the transformation done to us.

What skills are needed going forward?

Everyone I spoke to feels that when it comes to future skills, it really comes down to the application of our human brains and bodies to the work. Leautier talks about strategic impact – the crucial reframing of insights professionals for both client-siders and agency collaborators: “We need to be guardians of insight and become business partners.”

Meanwhile, Steadman underscores the enduring importance of storytelling, noting that researchers’ ability to interpret data and craft narratives from disparate data-sets remains a key value-add.   

Quinn-de Jong talks about how we will be managing hybrid AI agent-human teams in the near future and will need upskilling for this. She stresses that the real question isn’t whether to use AI, but where human intelligence adds value, and where to deliberately partner with AI. It requires conscious choices, not defaulting to AI doing the thinking for us. She sees a schism ahead: “I think there’s going to be a polarisation between the people who get to grips with it and take the opportunities it offers, and the people who reject it or use it in the superficially or uncritically.”

Rahul Manchanda – until recently leading AI transformation in insights at Pernod Ricard – reflects on the importance of change management skills. He notes that even in orchestrating complex systems change on a global scale, most of the time is spent not on the technology itself but on managing people and processes. He reflects on the positive opportunity for AI to support better marketing, which can lead to new ways of working: ”The predictive power of AI to anticipate consumer demand is strong.” To do this well requires significant investment in set-up and orchestration, he adds. 

Manchanda sees what we might call ‘soft’ human skills as being of crucial value in this new research world order: curiosity, empathy (for stakeholders and consumers alike), listening skills and cultural sensitivity. There is also a requirement for human critical thinking to translate into the business: “AI can create a lot of noise; you need to simplify it and make it relevant to help stakeholders make decisions.”

How ready are we, and how ready do we want to be?

Wilson is no rejecter of AI but reflects on the quality trade-offs it can bring. He notes that research is not just a process but a creative and intellectual act – and the learning is shaped through doing: ‘If you don’t do the moderating, you miss out on an opportunity to do the thinking’. He questions how rigorously crafted a discussion guide spat out of an LLM will be, how engaged will be the researcher who applies their brain as opposed to the one who defaults more actively to AI shortcuts, and how compelling AI-led deliverables are compared with real human voices drawn from, say, ethnography. He summarises his position: “I’m not anti-AI, but I’m pro-quality.”

There’s little room for the fear and rejection of AI in this sector, but it’s certainly not fashionable to push back against the tide. It reminds me of the Scottish referendum on independence. Hear me out on this. The framing of the ballot was ‘Should Scotland be an independent country?’. Those who disagreed were framed as rejecters. Those who voted ‘yes’ were voting in the positively coded direction and voting to embrace change.

I see the same here. It’s easy to position oneself as embracing change, as seizing opportunities to do things differently, faster and perhaps better. It’s harder to recognise those opportunities and apply a critical lens and hold on to what feels worth keeping from the pre-AI era, without carrying a risk of losing relevance and appeal. 

This is the tightrope many of us are walking as we adapt, and we should cut ourselves a little slack (while keeping that tightrope tight enough to be safe). Forgive the tortured analogy, but AI can tie us in knots. Someone stop me now.