A new Al can suss out the specifics. Given a snippet of dialogue between a chatbot and a human, the system - developed at McGill University in Montreal, Canada - predicts how convincingly human would rate the chatbot's response. In other words, it automates the Turing test. This could help build better virtual assistants. Today's chatbots are great for specific tasks - Amazon's voice-operated assistant Alexa can order a pizza, for example, or check the weather - but try asking if it's enjoying the weather
© 2001-2025 Fundación Dialnet · Todos los derechos reservados