Episode 86 of Robot Talk Podcast: Interview with Mario Di Castro (English).
About this episode:
In this episode of Robot Talk Podcast, we had the pleasure to talk with Mario Di Castro, the lead AI software engineer on the Google Duplex team.
Duplex is a conversational AI technology that can engage in human-like conversations and complete tasks over the phone.
In 2018, Google Duplex made headlines when it was demonstrated calling businesses and making appointments, sounding so human that many people were fooled.
In this episode, Mario shares his insights on the development of Duplex, the challenges of creating conversational AI, and the future of AI in customer service.
Key Points:
What is Google Duplex?
Google Duplex is a conversational AI technology that can engage in human-like conversations and complete tasks over the phone. It was developed by Google’s AI research team and was first demonstrated in 2018.
How does Google Duplex work?
Google Duplex uses a combination of natural language processing (NLP) and machine learning (ML) to understand human speech and generate natural-sounding responses. It also uses a knowledge base of information about the world to help it complete tasks, such as making appointments or ordering food.
What are the challenges of creating conversational AI?
Creating conversational AI is a challenging task. One of the biggest challenges is teaching the AI to understand human speech, which is often full of errors, hesitations, and interruptions. Another challenge is teaching the AI to generate natural-sounding responses that are appropriate to the context of the conversation.
What is the future of AI in customer service?
AI is expected to play a major role in customer service in the future. AI-powered chatbots and virtual assistants can provide 24/7 support, answer common questions, and even resolve complex issues. This can free up human customer service representatives to focus on more high-value tasks, such as building relationships with customers and providing personalized support.
Kind regards B. Guzman.
