Ask HN: Are people generally interested using LLMs for learning purposes?
A Hacker News user asks why using LLMs as interactive tutors to teach subjects like CUDA programming, with knowledge testing, never became a popular use case despite initial expectations about their vast knowledge.
Background
- The poster is asking the Hacker News community why AI-powered tutoring hasn't become a mainstream app, despite LLMs (large language models like ChatGPT) seeming perfect for interactive teaching.
- Hacker News (HN) is a tech-community forum run by Y Combinator; "Ask HN" posts are questions posed to the community. The audience is mostly software engineers, startup founders, and tech enthusiasts.
- LLMs ("large language models") are AIs trained on vast text data; they can answer questions, explain concepts, and generate exercises. Early hype imagined them replacing traditional search or teachers.
- The post reflects a real gap: while many people use ChatGPT to get quick answers or explain code, few use dedicated "AI tutor" apps for sustained, structured learning. Competitors like Khan Academy's Khanmigo or Duolingo's AI features exist but haven't exploded. Possible reasons the commenters discuss: LLMs hallucinate (make up facts), can't reliably assess understanding, and users prefer passive Q&A over structured lessons.