AI WireTechCrunch
Safeworld builds digital humans to test if robots are safe
Can a robot be trusted around real people? Safeworld is trying to answer that by building "digital humans," computer-generated stand-ins for people, to test generative AI robots before they interact with actual humans. The idea, according to TechCrunch, is to catch ways a robot might cause harm before it ever gets near someone in real life. As more companies use generative AI, AI systems that can create actions, text or decisions on their own, to power physical robots, questions about safety become more urgent. Safeworld's approach could give developers a way to check robot behavior in a safe, simulated setting first.
Words to know
- Generative AI
- — AI that can create new content or actions, like text, images or robot behavior, on its own
- Digital humans
- — Computer-generated figures made to look and act like real people, used for testing
- GenAI robots
- — Physical robots controlled or guided by generative AI systems
Summary written from TechCrunch's headline and teaser