Fireside Chat, Interview
Indexing Humanity: A Fireside With the Simulation Company | Simile | RAISE Summit 2026
- Jun plans to develop foundational models capable of predicting individual subpopulation behaviors and attitudes with 85% of the accuracy of human self-replication, transitioning the industry through four stages: individual static simulation, individual dynamic simulation, multi-agent simulation, and society-wide simulation on a rich world model.
- Enterprise partners are expected to replace major consulting firms by running entire operations in simulation to alter operational cadence, potentially resolving current constraints where companies execute only one-tenth of desired studies due to human resource bottlenecks.
- The technology is positioned to evolve from a nascent stage comparable to early Generative AI eras, with Simile's generalizable architecture expanding horizontally beyond market research into verticals such as finance, CPG, retail, and telcos to model overlapping target populations.
- Predictions include a shift to an agent-driven age where decision-making rates become "unimaginable in the past," content creation costs drop to zero, and new use cases emerge for product development and earnings call simulations.
- Critical future challenges involve the growing gap between agent outputs and human values, requiring simulation to inject human voices into high-level government and product decision-making, while adoption relies on data transparency and consent rather than purely on prediction accuracy.
- Significant risks include the potential for simulation technology to cause serious damage and clear cases of misuse, prompting a strong emphasis on upholding informed consent and user autonomy principles.