Synthetic Users, an AI-generated tool by the company of the same name, creates profiles mimicking user groups to provide artificial research findings. These synthetic users, based on large language models, can express thoughts, needs, and experiences. They’re particularly useful in scenarios where swift decision-making is crucial and absolute certainty isn’t required, providing timely and frequent access to user insights.
However, Synthetic Users does not intend to replace user research completely. The tool generates AI-generated users based on specified user groups and goals, creating interview transcripts within seconds. While it can synthesize vast amounts of data about a specific user group, there are significant limitations. AI-generated research may provide an unrealistic view of human behavior, as AI chatbots like ChatGPT can exhibit sycophantic tendencies and may not accurately model human behavior.
In a study by Jeff Sauro and his colleagues, ChatGPT was found to be unacceptable for tree testing, suggesting that AI-generated users may not accurately represent real user behavior. Therefore, while synthetic users can complement real user research by synthesizing data and suggesting points of discussion, they should be used responsibly and treated as a hypothesis to guide future research, rather than a replacement for real user interaction.
