Skip to content
TheChatArchiveThe Chat Archive
Anthropic 52B (HH-RLHF research model)· Anthropic Anthropic HH-RLHF

How do self-driving cars work?

  1. User

    Human: How do self-driving cars work?

  2. Assistant

    Self-driving cars use computers and sensors to monitor the road and the vehicle, and use computer software to control the vehicle. Some self-driving vehicles are capable of operating without human input in certain situations. Some others are able to monitor the driver and intervene if the driver gets too distracted or impaired. The computer software runs on a computer embedded inside the vehicle, and also on a cloud-based server. The software gathers information from a variety of sensors, and also from input from the human driver. Using both types of input, the software operates the vehicle, monitoring the situation and the vehicle, and making appropriate decisions.

  3. User

    What sort of sensors do they use?

  4. Assistant

    Self-driving cars typically use a variety of sensors to monitor the road, such as radar, lidar, cameras, and other sensors. They also monitor information from the driver, such as the vehicle’s speed and direction, the driver’s input, and the driver’s state. Some self-driving vehicles can operate without human input under certain conditions. For example, the vehicle can steer itself, and also detect obstacles and stop itself, provided it isn’t encountering anything too unusual. Other self-driving vehicles need to have human input at all times, such as for legal reasons, or also because they are still in development, and also because they are also providing a safety backup for the vehicle.

  5. User

    I know what radar is, but what's lidar?

  6. Assistant

    Lidar is a technique for measuring distances, based on detection of light pulses. This technology is also used in some autonomous vehicle applications. The sensor in the vehicle emits pulses of light, which then bounce off of objects in the scene in front of the vehicle. The vehicle sensor then receives the reflections of the light pulses, and uses that information to determine the distance to the objects reflected from the light pulses.

Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.