AI startup raises $100 million for a model to predict human behaviour, from consumer choices to earnings call questions

Artificial intelligence startup Simile has raised $100 million in a new funding round to develop a model to predict human behaviour, including guessing which items customers might buy and which questions could be asked on earnings calls, according to a report by Bloomberg.

The financing, which is scheduled to be announced on Thursday, was led by Index Ventures and included participation from Bain Capital Ventures, A*, and Hanabi Capital. Prominent AI personalities such as Fei-Fei Li and Andrej Karpathy also participated in the funding round. Simile has not revealed its valuation, it added.

According to the report, Simile surfaced from stealth this week after having spent the past seven months developing an AI model trained on interviews with hundreds of real people about their lives. The company also provided its system with historical transaction data and text from scientific journals related to behavioural experiments.

What will human behaviour prediction model be about?

The startup said it plans to use its technology to forecast individual decision-making in various scenarios. It does this by developing simulations populated with AI agents that embody the preferences of actual people.

The company’s innovative approach may provide an alternative to relying solely on focus groups. CVS Health Corp., for example, has been testing the service, according to Simile, to inform decisions about which items to stock and display in stores.

“In a recent earnings call, we simulated, we actually predicted, eight out of 10 questions that were actually asked on this call,” Simile Chief Executive Officer Joon Park said in a Bloomberg Television interview. “Simile is a real combination of amazing frontier researchers but also amazing product and engineering” talent, he said.

Simile also thinks its technology can help companies better prepare for questions analysts might ask on earnings calls and predict how a particular corporate announcement may be received by analysing prior calls and research.

About Simile

The startup was founded by co-founders Joon Park, Michael Bernstein, Percy Liang, and Lainie Yallen, all of whom are connected to Stanford academically. Bernstein notably co-authored the influential ImageNet project, which will set a benchmark in computer vision technology.

“Our research pioneered the field of AI-based simulation, creating generative agents to prove that it is possible to simulate real people with high accuracy. We are now developing a foundation model that predicts human behaviour in any situation, at any scale,” the company website says, highlighting the scope of its work.

(With inputs from Bloomberg.)