AI: anecdotes, experiences, and hallucinations
Manage episode 509034517 series 3682007
The AI team take a deep dive into a talk that shares experiences using AI as a smart search engine, how it can provide new insights when using AI as a search engine, and assist with writing textbooks, conference and journal papers, in the classroom, in systems engineering tools, creating images and graphics, creating and debugging software, marketing, self-reflection, creating podcasts and presentations. The talk also covers some ethics and the potential for scams. The talk discusses the dialog and iterative active listening needed for effective communication. It warns that the AI is a great sycophant, apologizes very well, and tends to hallucinate, so that we need to ‘trust but verify’ Al the information it produces. The talk concludes that AI is a very useful tool, but has to be treated as an apprentice who has access to a lot of information but needs to be guided and constantly checked. The talk was given at the BCS SQM XXXIII 2025 International Quality Conference on 19 September 2025. The YouTube URL of the recording is https://youtu.be/0LPqiMA_j-0
10 episodes