Tesla's Project Rodeo is an ambitious safety testing initiative designed to push the limits of its self-driving software, relying on a specialized group of test drivers.

The company markets its autonomous driving system as 'supervised self-driving' (FSD), which is currently in use by customers while also operating an internal testing fleet.

While none of the interviewed test drivers reported being in accidents, they described numerous near-collision experiences, including close calls with pedestrians.

This approach poses significant risks, as test drivers sometimes allow the car to roll through red lights or swerve into other lanes, exposing the public to potential danger.

Experts have highlighted the need for clearer regulations and oversight in autonomous vehicle testing, comparing the current landscape to a 'Wild West' environment.

Test drivers are trained to delay intervention during tests, even in dangerous situations, to maximize data collection on the software's performance.

A controversial aspect of Project Rodeo involves 'critical-intervention' test drivers who are pressured to postpone interventions until the last moment, often in risky scenarios.

The pursuit of data collection has created a high-pressure atmosphere, where many involved feel like unwitting participants in a dangerous experiment.

Philip Koopman, an expert in autonomous driving, criticized Tesla's methods as 'irresponsible' due to the risks posed to uninformed road users.

In contrast to Tesla, companies like Waymo and GM's Cruise conduct their testing in controlled environments, emphasizing safety and quicker driver interventions.

Despite conducting extensive simulations, Tesla's reliance on real-world testing raises concerns about the potential dangers involved in their data collection methods.

Tesla's safety program aims to gather real-world data from high-risk scenarios to enhance its self-driving AI models, similar to strategies employed by other autonomous vehicle companies.