Oscar Carlsson on generative AI in the workplace

Oscar Carlsson, Chief Innovation Officer at Cint recently joined our podcast. Cint is a market research technology company that provides innovative solutions for businesses to conduct research quickly and efficiently. Oscar oversees innovation and new business areas, as well as presales and the sales engineering team, and is involved in merger and acquisition activities and thought leadership. He also focuses on delivering the highest data quality and mitigating fraud in the industry. Below is an edited transcript of our discussion. 

Michael Howard: So Oscar, what's the most significant thing you're going to tell us today?

Oscar Carlsson:
So I think the main thing is because AI is such a hype and buzzword out there, Cint set out to actually do some research behind the usage of AI and how AI plays into the workplace. So happy to share some findings from that research, which might in some cases be a bit eye-opening and in some cases just validate what we believe already. And then happy to share more about actual use cases where we use AI already at Cint and some of our future plans and where we think the market research industry is heading in the use of AI.

Michael Howard: Looking forward to that. So I mean, the conversation around AI certainly is ever-evolving. There's always something being released today or tomorrow. So it's going to be an exciting discussion. To set the scene though, can you tell us a little bit about Cint's journey into AI?

Oscar Carlsson: Of course. Cint has always been the global leader in digital insights, and then what we do and where we sit in the marketplace is that we say that we set out to help feed the world's curiosity. And then we do this by running services, but really where we fit in is in the supply chain where we optimize the procurement of a panel for market research studies, and that's where we first came into AI. It was to really use AI in the most efficient way to find the right panel and the right respondent for your study, where we see the highest conversion rate without introducing any sort of bias and then AI. We've seen more and more use cases, especially in finding patterns in terms of bad data quality and where that is coming from to really help eliminate the fraud and the bad actors in the space by essentially predicting if incoming respondents would provide bad data or not, based upon the billions and billions of data points that we run through the platforms.

Michael Howard : It's true. It is a really interesting space, and we'll get onto some of this shortly. So you mentioned the study earlier on AI in the workplace. Can you tell us a bit more about this and some of the findings in it?

Oscar Carlsson: Yeah, of course. So we set out to find people who were full-time employees. We did this in the U.S., U.K., and Australia. And we asked if people were using AI at work and then how they were using it. And also if their colleagues were aware of it, if the company was aware of it. Did they get any guidance or were there any policies in place by the companies? And really, we named the study of this, that AI is being a double-edged sword and why. The main takeaways from the study were maybe not surprising that a lot of people use AI, they see the power with it, but also the danger with it. And one of the findings was that in the U.S. over half of the employees think their company needs to have an AI policy in place, while only one-fifth of the companies actually do have a policy in place. So it's sort of out there in the wild currently.

Some of the other things that we learned from the study is that 20 percent of companies are actually encouraging people to use AI because it's really heightening productivity for the people working in the company. But only 20 percent of the company is encouraged, but 65 percent of the managers encourage it. So it's more of the people's managers who encourage it than the actual companies behind it, which proves that companies are still on the back foot of guiding their employees on how to use it. They're not really sure about the implications. So there hasn't been any official policies or guidance yet. But what we see from the data is that if people’s managers already see the value, it's about to come very soon.

Michael Howard: I've worked in tech for quite a while now. And one thing that I've seen is once someone is used to working using tech as part of their productivity and their job. They find it so difficult to go and work in a company that doesn't use it. So it's actually a huge turnoff and dissuasion or barrier for people to join a company if they're not up to speed with it.

Oscar Carlsson: Yeah, I think it's just such a major shift now in the workforce with the pandemic being there. People are used to working remote or hybrid and then with AI coming in as well. And especially now the younger generation being, of course, initially digital natives but then also using AI. And then the way I see, especially generative AI is that, of course, when we were all young, we had to do math by hand, and then we got the calculator. That was the first step, which created a lot of efficiency, but we had to know the math behind it. And the next big revolution to me was, of course, the internet came, but really when Google came, and we started using Google searches. So you don't have to retain all the information, but you knew how to find it. And you used criticized it when reading all the sources, right? And I think that's the way we should see generative AI as well. It's sort of Adding the same efficiency as the calculator and Google - having both the sources of all the information-  but also can help you produce content, but you still have to apply critical thinking because it's still in its infancy.

Michael Howard : That's right. You mentioned the use of calculators and the timeline you just gave brought up a memory of mine, the calculator with the solar panel on it. So you didn't have to ever put a new battery. That was amazing.

Oscar Carlsso : Yeah, technology revolutions. It's exciting, but it's also, of course, you have to think about the power and what's behind it.

Michael Howard: So, and I think we're coming up in a really interesting age where we are actually realizing where it's coming from. And I think in probably 20 years from now, people have no idea how AI actually works and where it came from. It's just a natural part of life.

Oscar Carlsson : Exactly. And I think the sort of generational shift, which is happening now, is that the younger generation that enters the workforce, they will have that generative AI as part of their life, and it will help them do things more efficiently and will add a lot of efficiency to that. And just like we have calculators in Excel for the older generation, it will be sort of an equivalent. So that's, of course, really, really exciting.

Michael Howard: So Oscar, thinking about your research and your findings, what do you think this means for companies in general and its significance for the market research space in particular?

Oscar Carlsson: Yeah, I think it's a really good question. So the new technology always opens up new possibilities, but the lack of regulation is still a concern, especially in the market research space, which is all about privacy and data usage. So the main takeaway from the research, I think, for companies in general, but especially in the market research space, is that data privacy is crucial. You need to understand where your data is stored, how it's managed, and what the options are. And then one of the most exciting things, of course, for a market researcher is that data accuracy is crucial, and generative AI will help you identify bad data quality and eliminate fraud. So the main takeaway for companies, but especially in the market research space, is that companies need to navigate this reality and support the responsible integration of AI tools into their insights toolkit.

Michael Howard: And then there's another element to this whole AI discussion that doesn't often get talked about that much, and that's explainability. How organizations need to be able to advise how their AI has produced a particular output or outcome. How do you see this playing out with Gen AI in the future?

Oscar Carlsson: That's a really good point. So explainability is, of course, challenging since AI is a black box. And we've seen a lot of regulation, for example, in Europe now coming in where you have to explain why certain things happened. And the main thing for companies, especially in the market research space, is that transparency is key. You need to be transparent with your customers on how AI operates, how data is used, what the sources of the data are. And then you have to be critical to the results as well and explain the process to your customers. And that, of course, is very difficult when you have a fully automated system. And that's one of the challenges that we see and that we're working on is that how can you explain why you've made certain predictions, why you've put certain data into the system? And that is a major challenge for the industry to solve.

Michael Howard: Yeah, and then if you do have that internal, that primary data access, it could be market research data or sales transactional data. You can actually use that to analyze historical data and even test it against your historical results and findings as well. And it's a good way to test the efficacy of it all as well.

Oscar Carlsson: Absolutely. Benchmarking and testing against historical data is crucial to ensure the accuracy and efficacy of AI-driven insights.

Michael Howard : Yeah. And what do you feel market researchers aren't talking about or thinking about when it comes to AI or market research software to that extent as well?

Oscar Carlsson : Yeah, that's a really good question. So I think many market researchers might not be fully aware of the power of AI. And of course, it's essential for companies to advise and train their staff, set an AI policy, and really create small innovation ops to test new tools and see how they can be integrated into their existing workflow. And then, of course, when you bring new tools into the organization, it's crucial for employees to embrace AI. But at the same time, be cautious with sensitive data, especially in the market research space where privacy and data usage is very, very important. And then just like any tool, you need to be critical of the results as well and really test and see if it's actually adding value to your process. And especially when it comes to generative AI, it's a very powerful tool, but it's also in its infancy, and you need to test it carefully and see how it actually can help you in your specific use case.

Michael Howard : So Oscar, is there anything else you'd like to add to help summarize our discussion today?

Oscar Carlsson : Yeah, I think for the main takeaways is that companies should advise and train their staff, set an AI policy, and really create these small innovation ops to test new tools. I think employees should embrace generative AI and use it for productivity. At the same time, be cautious with sensitive data, especially in the market research space. And then be critical of the results when you test out generative AI in the market research space and really see if it adds value to your process. And then, of course, for companies that actually effectively utilize AI, they will have a competitive advantage in the market.

Michael Howard: Well, thank you so much for joining us. Really excited to have you on the show and to hear how you at Cint are helping to make market researchers' lives easier and consumers' lives better. Thank you for having me. It's been fun talking about this, and I'm really intrigued to see what everyone will do with AI in the next 6 to 12 months and where we are a year from now.

Oscar Carlsson : Exactly. We'll have to catch up and have a review to see how we've gone and where we're right or did we miss the mark completely?

Michael Howard : Awesome. Thank you very much.

Oscar Carlsson : Thank you.