Chatbots and the future of online learning
summary
Duolingo builds conversation practice as scripted chatbots inside the app, unlocked at higher levels, improving by collecting user interactions and applying machine learning to rank and refine scripts. A surprising early finding is learners often “don’t know what to say,” so a help me reply hint supplies usable responses. One playable scenario, Pizza Time, branches based on replies but remains mostly script-driven; retention and learning metrics trend positive as models open up.
AI-written summary · about
chapters (16)
- 0:00Introduction to chatbots
- 1:03Duolingo chatbot overview
- 3:13Why Duolingo added chatbots
- 4:36Dynamic conversation challenges
- 5:58Chatbot learning and progression
- 7:03Unexpected user behavior data
- 8:25Chatbots in daily life
- 9:31Measuring chatbot success
- 10:12Timeline for advanced chatbots
- 11:13Education advantages of chatbots
- 12:17Defining chatbots clearly
- 13:19Future of chatbot adoption
- 14:46Human-like chatbot risks
- 15:51Chatbot customization trends
- 16:32Open sourcing chatbot tech
- 17:16Career paths in chatbots
related talks
10mEnglishBot: An AI-Powered Conversational System for Second Language Learning
35mDesigning Experiences for Chatbots
01h 04mBuilding Expertise Into Generated Conversations
50mThe Art of Coding a Conversation: Designing Bots
32mChatBot Learning Assistant with Large Language Models
52mAI-Powered Learning Pathways & Skills Development
20mAI in Android Development at Duolingo
24mChatbots — Our Conversational Partners Today and Tomorrow