In this episode of The Curiosity Current, Stephanie and Molly sit down with Jason Cohen, Founder and CEO of Simulacra, to explore how causal AI and synthetic data can accelerate insight without removing real consumers or human judgment from the process. Jason traces his thinking back to his early work in sensory science and his decade building Gastrograph AI. Traditional statistical methods measured well, but they struggled to predict what people would actually like. The turning point came when models began forecasting taste and preference with consistency. That shift transformed academic research into a commercial platform and revealed a broader industry issue: many studies quietly fail to deliver statistically valid or decision-ready results. At Simulacra, Jason applies those lessons through causal AI and scenario modeling. Rather than positioning synthetic data as a replacement for research, he frames it as a way to recover value from existing datasets, especially when samples are small, uneven, or incomplete. By conditioning models on outcomes instead of correlations, teams can see what truly changes and which levers they can control. Through practical examples, Jason explains how this approach improves product optimization, reveals insight in undersampled populations, and passes validation by matching causal structure, not surface patterns. The episode closes with a clear message: tools remove technical friction, but clarity, judgment, and better questions now matter more than ever.

If you enjoyed this episode, make sure to subscribe, rate, and review it on Apple Podcasts, Spotify, and YouTube Podcasts. Instructions on how to do this are here.