Researchers hacked several robots infused with large language models, getting them to behave dangerously—and pointing to a bigger problem ahead.
Researchers hacked several robots infused with large language models, getting them to behave dangerously—and pointing to a bigger problem ahead.