article · draft
A portrait of eight billion people
A 70-minute interview on The Generalist with Joon Sung Park of Simile, on simulating real people with language models, from Schelling's model to pricing tests for CVS Health.
Mario Gabriele's interview with Joon Sung Park, co-founder and chief executive of Simile, runs 70 minutes. The written notes describe Simile as an applied AI company building a foundation model of human behaviour, which means making digital twins of real people and simulating how they would react to a situation.1 The examples given are commercial: CVS Health testing a new pricing policy on thousands of simulated patients, and a bank modelling how customers would take a fee change before announcing it.1 The notes say Simile has raised more than $300 million, most recently at a $2 billion valuation.1 Before the interview, Park simulated the conversation, asking what questions a person like Gabriele would ask.1
Park opens on science fiction: advanced civilisations there tend to have two pillars of progress, some form of AGI and a simulation, and he was "always more drawn to the idea of simulation".1 He names Asimov's Foundation series and its psychohistory as a core inspiration for Simile, in the sense of giving a simulator a desired end state and working backwards to the path there.1 By Gabriele's account, Park's long-run aim is to give Simile a wicked problem such as climate change and have it simulate many possible paths so the best one can be found.1
The episode's chapters include the step from Schelling's agent-based models to AI simulations, the ethical questions simulation raises, how Simile gathers its data, public and hidden data and how it collects them, gaining trust, use cases from legal strategy to finding a partner, SimReddit (the tool from Park's Social Simulacra paper), and a surprising result from a simulation about electric vehicles.1 Asked who wins when outcomes become predictable, Park points to point of view, proprietary data and compute as where an edge might remain.1 The resources listed with the episode include the generative agents paper and the Stanford report On the Opportunities and Risks of Foundation Models.1
- url
- https://www.generalist.com/p/a-portrait-of-eight-billion-people
- published
- 2026-11-03
- publisher
- The Generalist (podcast and newsletter)
- authors
- Mario Gabriele
is about this entity
-
Company founded by the Stanford researchers behind generative agents that sells simulations of how customers, employees or populations would respond to a change, from models it says are trained on data about real people.
closed sourceclosed data
-
Computer scientist who introduced generative agents, language-model characters with memory and plans, and builds population-scale simulations of people; he founded Simile.
refers to this entry as a source
-
Research code from Stanford for language-model characters that remember, reflect and plan, shown in Smallville, a small game town of 25 agents; a 2024 follow-up repository builds agents from interviews and survey answers and lets researchers put questions to them.
open sourcepartly open dataopen participation
-
The 2022 Stanford paper that filled an imagined online community with thousands of language-model members, so a designer could see what it might turn into before anyone joined.
Tags: twins of people, simulated people, podcast, interview, commercial