San Francisco startup Mirror Particle is set to launch an AI engine designed to predict human behaviour, arguing that existing methods relying on large language models (LLMs) for role-playing fall short. The company, founded two years ago, believes LLMs are fundamentally limited in understanding human behaviour because they primarily model written language, rather than visual perception, spatial reasoning, and social intelligence.
Mirror Particle's co-founder and CEO, Abhivyakti Ahuja, stated that their approach involves building a 'world model' from scratch. This model aims to simulate why humans act as they do and how their behaviour evolves over time, focusing on 'revealed behaviour' rather than self-reported survey answers.
The startup uses a combination of data, including client customer data, current events, pop culture, and social media, to model demographic segments. Mirror Particle's initial strategy targets market research and brand and product strategy, providing insights into consumer motivations and constraints. The company has already secured an angel round of funding and is reportedly close to closing its first venture round. Mirror Particle is also scheduled to compete in TechCrunch's Startup Battlefield next week.