THE LIVING ARCHIVE

From punched cards to autonomous collaborators: the ideas, people, languages and systems that taught machines to follow—and increasingly interpret—human intent.

PAST / PRESENT / NEXT

Collection

Tag: adversarial-testing

Microsoft Tay: How a Chatbot Learned the Worst Lessons of Twitter in Less Than a Day

Microsoft's Tay experiment showed how a public conversational agent could be manipulated through coordinated interaction, exposing weaknesses in abuse testing, response controls, and deployment boundaries.

↗