Everyday AI Podcast – An AI and ChatGPT Podcast EP 158: The ChatGPT Mistake You Don’t Know You’re Making
Dec 5, 2023
Discover common mistakes that lead to hallucinations with ChatGPT and learn how to avoid them. The discussion includes critical insights into recent AI developments and industry collaborations, emphasizing safety concerns around AI technologies like Tesla's self-driving system. Explore knowledge retention challenges with ChatGPT and gain tips for better interaction, including updates on document uploads and custom configurations. Enhance your AI experience while navigating the complexities of generative AI.
AI Snips
Chapters
Transcript
Episode notes
ChatGPT Quality Concerns
- Many users report ChatGPT's quality has decreased, with incorrect information and hallucinations.
- Jordan Wilson observes this might be due to users misunderstanding how document uploads and knowledge retention work.
ChatGPT Memory Limit
- Remember ChatGPT has a memory limit (currently 32K tokens, not 128K as advertised).
- Test features yourself; don't solely rely on company announcements.
Document Referencing Example
- Jordan uploaded a writing sample PDF with hidden information (favorite color).
- ChatGPT couldn't answer "What is my favorite color?" without being explicitly told to reference the document.