Founded in 2024, Artificial Societies develops AI-based digital simulations of ‘societies’ – groups of AI personas. The company is backed by Y Combinator and Kindred Capital.

Users of the simulator, ‘AS’, can experiment with different product or marketing messages, social posts and concepts before they launch them. 

With each simulation, the platform generates 10 alternative content variations, and marketers can choose tools focused on brand growth, raising capital or general product-market fit.

Artificial Societies uses data collected at an individual level from channels including LinkedIn and X. The company’s behavioural analysis system identified patterns in preferences and motivations to create a database of around 500,000 AI personas underpinning the simulation tool.

James He, chief executive and co-founder of Artificial Societies, said the approach offered “open access to understand markets, audience sentiment, and to drive brand engagement”.

The company conducted a comparison study that found the simulation tool more accurately predicted how LinkedIn posts would perform, compared with subject-matter experts and large language models (LLMs).

Commenting on Artificial Societies, Jared Friedman, managing director, software and group partner at Y Combinator, said: “Instead of replacing an existing budget, they have built a product that will expand what it means to do market research. Their new AS product is an innovative way to let founders and marketing folks experiment with their messaging in a proven simulation before taking the bet.”