Designing for humanlike interaction
summary
Team frames concrete design principles for conversational AI used on Cortana and a personality-chat bot framework: digital transparency (always disclose nonhuman), reject the Turing test (don’t try to impersonate people), and a canned refusal strategy — detect harassment or sensitive confessions and reply with a neutral acknowledgment like there’s nothing interesting here rather than humor or shaming. Goals: reduce abuse amplification, improve inclusivity; deployed in bots and reported to cut escalation in production.
AI-written summary · about
chapters (5)
related talks
45mDesigning meaningful Human-AI Interactions
57mDesigning Human-AI Experience
28mHuman Computer Interaction: The UX of AI
35mDesigning Experiences for Chatbots
36mStrategies for Human-AI Interaction and Collaboration
29mAvoiding the Chatbot Trap Designing AI Interfaces That Unlock Real Business Value
44mNavigating the Hurdles of Conversation Design
20mEmbracing the Voice Revolution