Laplace Intelligence red-teams robots and self-driving cars before they hit the real world

Howie Sun, Laplace’s CEO and co-founder, told TechCrunch that the OpenAI-Hugging Face incident showed AI agents can be unpredictable. “Imagine if something like that happened in the physical world,” Sun said. “AI agents are moving from computers into the real world, and we don’t really want physical AI to fail in the real world.” He said Laplace aims to help companies discover and fix problems before deployment and, over time, build a shared foundation of evidence that customers, insurers, and regulators can trust.

Sun and co-founder and chief scientist Henry Liu started Laplace a year ago out of the University of Michigan. Sun has spent the past eight years in academia researching how to challenge autonomous driving AI in simulation. In 2023, he and Liu, his then-PhD advisor, published a paper on the subject that appeared on the cover of Nature. Sun said that led to calls from autonomous vehicle companies interested in implementing their ideas, and he decided to turn the research into a product useful in the real world.

The company takes its name from Laplace’s demon, an 1814 thought experiment by French scholar Pierre-Simon Laplace that imagined predicting the future with total precision if one knew the exact state of everything in the universe. “We are essentially using simulation to predict the real world’s future,” Sun said.

At Laplace’s core is a high-fidelity simulation reconstructed from the real world, paired with an AI agent that constantly probes for weaknesses and failure points. The goal is to ensure physical AI—whether a robot, autonomous vehicle, or smart city solution—is safe in simulation before deployment. The simulation environment is a post-trained version of Nvidia’s Cosmos, an AI world simulation platform, grounded in the real world, allowing Laplace to reconstruct settings from an Arizona highway to a Paris pedestrian crossing.

Simulations are generated by blending a mesh of different real-world information. A Google Street View image could serve as a reference from which an entire street is generated. Rather than generating something random, the simulator combines real-world map data to recreate an environment true to scale. Developers can use language prompts to change the setting in real time, such as adding afternoon traffic. Sun said Laplace pulls real-time traffic and congestion information from navigation providers like TomTom.

Within the simulations, not only the environment is simulated, but also the behavior of other actors. In a demo, Sun directed the test agent vehicle to swerve into oncoming traffic. The car it would have hit—or driven through, as is common among some generative world models—moved out of the way to avoid a collision. “It’s truly an interactive simulation environment,” Sun said. “Everyone is reacting to everything. … This is more like a world model with interactive participants.”

While the simulated environment is important, Sun said the unique selling point is the AI agents Laplace is developing to automatically design corner cases, moving humans further out of the loop. “They can design from the news, from police reports on collisions—all kinds of information to actively challenge the autonomous vehicles and other physical AI agents,” he said.

Today, autonomous vehicle developers mostly train AI to handle edge cases using recorded footage that they turn into a simulation, or by relying on humans to come up with new ideas and design them. TechCrunch reported that many AV companies have begun embracing AI to create edge cases for training, citing self-driving truck company Waabi, but Sun said the need for this technology will become more apparent when generalized robots are placed in the real world. “There’s less data on generalized robotics,” Sun said. “No one knows what the edge cases will be in the real world.”

Despite Sun and Liu coming from academia without Silicon Valley tech company pedigrees, they have started generating revenue and turning a modest profit. TechCrunch reported that most of the money is coming from autonomous driving use cases.

Editor's Summary Laplace Intelligence is developing automated red-teaming for robots and physical AI, using high-fidelity simulation and AI agents that generate corner cases to uncover failures before deployment. The Ann Arbor startup, founded by former University of Michigan researchers, is competing in TechCrunch Disrupt’s Startup Battlefield and has begun generating revenue, mostly from autonomous driving use cases.