Dr. Olivier Toubia, Professor at Columbia Business School, joined Askable in this webinar to unpack his team's 19 studies on AI personas, the danger of confidently wrong data, and where real human evidence remains irreplaceable.

Product and design teams are under immense pressure to ship faster than ever. Because of this, the pitch of synthetic users and digital twins is enticing: instant, scalable customer data without the cost or friction of scheduling real participants makes sense to try. But can an AI model actually replace human discovery?

In this exclusive webinar, we are bridging our rigorous standard for verified human evidence with Dr. Olivier Toubia’s groundbreaking empirical data. We will cut through the noise to explore where synthetic tools provide valid early signals and where they critically distort the messy, unpredictable reality of real users.

What would it take to hear from real people as quickly as you can ask an AI?

What you'll learn:

• Where digital twins reflect people and where they fall short, including why they can be overly rational or miss the diversity of human responses.

• How to evaluate a twin for a specific use case, from choosing meaningful benchmarks to testing its predictions against real people.

• How different sources of data shape a twin’s usefulness, and where it can serve as a research thought partner alongside (not in place of) human evidence.

About our guest:
Olivier Toubia is the Glaubinger Professor of Business at Columbia Business School. His research focuses primarily on innovation, customer insights, and creative industries. Specifically, he combines methods from social sciences and data science in order to study human processes such as motivation, choice, and creativity.