Across the two days it was clear there was a major key theme emerging: the use of AI and everyone’s most spoken about topic:ChatGPT. The below discusses the implications and the practical uses within market research (are we- humans- going to be replaced?) AI technology has taken a giant leap forward with the launch of OpenAI’s ChatGPT and other large language models (LLMs). Although we can’t yet predict the way in which this will change the world or the way we work, from what we’ve seen, we do know it has huge potential to disrupt our current ways to working. Like most industries, market research will not be protected from these changes, and we need to be alert to the challenges it will pose. However, in many ways market research will be on the frontline of finding applications for this new technology- it has the potential to optimize the way we work or create entirely new approaches to research. It is therefore not surprising that ChatGPT was a popular topic of discussion at the recent Quirks London event. Potential Watchpoint: Data validity The immediate concern for our sector is the validity of our open end survey responses. There is the potential that people could use Large Language Models (LLMs) to impersonate healthcare professionals (HCPs) or patients. LLMs could also be used to automate survey responses en mass meaning their existence could become a real problem, if not monitored. The trick to addressing this issue is us- and by us, we mean the imperfections that make us human. LLMs are programmed to be ‘perfect’, providing grammatically perfect answers, in a perfectly formed structure. Our respondents aren’t like that, their responses are shorter with sometimes poor punctuation, misspelled words and capitalization (e.g., EFFICACY MOST IMPORTANT 1L). These naturally occurring human imperfections are a good thing- they are traits we can use to spot potential fraudulent responses. Further to this, LLMs are ultimately a program, created with morality and self awareness. They won’t respond to subjective questioning. Therefore, another method of protection may be to modify our lines of questioning in subtle ways (i.e., discuss attitudes, feelings, and perceptions around a topic). The first step to mitigate against these challenges is awareness- HRW are working closely with our expert partners to build more sophisticated ways to detect machine-generated responses, it’s essential that we stay ahead of the curve. Potential Benefits: Research Optimization As well as being aware of the challenges, healthcare research companies must also be proactive in integrating new technologies to improve the way we work. LLMs are already proven to be adept at admin tasks but testing has shown that the expertise and insight of a researcher is, and will likely always be, required. During the conference ChatGPT was asked to ‘write a discussion guide to explore a topic’- and yes, a well-formed discussion guide was sent back in response. However, anything that’s generated needs to be curated- there is still a need for skilled researchers to tweak outputs to ensure the more nuanced research objectives are addressed- we can’t take everything at face value. LLMs will be a new tool in our research team’s arsenal rather than a replacement. Machines can generate responses, but they cannot replace the ‘human factor’: the insights and expertise that come from years of experience working in different therapy areas and across different methodologies. The best research comes from a combination of advanced technology and human involvement, with each playing a unique and complementary role. HRW are making it a priority to monitor the developments in the AI space and constantly analyse the impacts and opportunities. Our Innovation team are exploring ways to integrate generative AI tools into our methodologies where necessary, and we closely follow latest regulations and compliance concerns as this topic develops. We also have research on the horizon that tests the efficacy and utility of generative AI, and we hope to share more on this topic soon. Please reach out to HRW Innovation at firstname.lastname@example.org if you have any questions or would like to hear more about our time at the QUIRKs Conference. By Darren Vircavs and Abigail Graham Apply Now!