Simile Raises $200M to Create AI ‘Twins’ of Real Consumers
Simile, a year-old Palo Alto startup, has secured $200 million at a $2 billion valuation to develop "agentic twins," AI-generated representations of real consumers that companies can survey instead of traditional market research.
What’s the Deal?
- The Pitch: Simile claims to predict consumer behavior by simulating it, offering faster and cheaper testing for new products or brand launches.
- Clients: CVS Health, Deloitte, and Wealthfront are among its early adopters, utilizing 400,000 agentic twins for various tasks.
- Technology: Founded by Joon Sung Park and co-created by Percy Liang, a pioneer in AI research, Simile’s model was trained on data from interviews with 1,000 individuals, supplemented by input from millions more via Gallup.
- Accuracy Claims: The startup boasts 85–99% accuracy rates, but these figures are disputed and based solely on their own evaluation.
The Controversy
The concept of AI stand-ins for human consumers is not without debate. Critics question the validity of such simulations in replicating real-world behavior. Simile’s goal to model all eight billion people globally adds to the skepticism.
"If you’re able to simulate the world, you can basically test out countless interventions," Joon Sung Park, CEO of Simile, passionately stated, emphasizing what they call "a window into humanity."
Despite the challenges and questions, Simile’s rapid growth—from $100 million six months ago to $200 million today—indicates a strong market interest in this controversial yet potentially game-changing technology.