AI robustness

When we build AI systems - care should be taken to test its robustness. A decentralized group of safe streets activists in San Francisco realized they can disable Cruise and Waymo robotaxis by placing a traffic cone on a vehicle’s hood, and they’re encouraging others to do it, too.

This may be a prank, but the larger point remains - how to ensure that AI systems are robust. In the paper titled "Robust Physical-World Attacks on Deep Learning Visual Classification" in 2018 researchers did multiple experiments to test adversarial attack on these vision models.