Why operate in meatspace, talking to humans, when you can interview sims? As LLMs become more popular and accessible, the risks that come when you avoid talking to humans grows exponentially.
Everyone knows they need to talk to customers - but nobody wants to. Maybe it costs too much, maybe it's too slow, maybe it's just too hard. But you can't just simulate a survey with LLMs and expect to learn what you need to learn to innovate and grow.
So let's talk about it! When are LLMs helpful in conducting customer research? When are they not? What do you gain from using them, and what do you lose? What should you watch out for in sourcing, authenticating and validating data? And why is human connection more important than ever in a new era of synthetic feedback?