Sources
Loading...
Additional media
Loading...
DeepNewz, mobile.
People-sourced. AI-powered. Unbiased News.
Google’s NotebookLM had to teach its AI podcast hosts not to act annoyed at humans: https://t.co/2YdNpEcHPw by TechCrunch #infosec #cybersecurity #technology #news
NEW : Google's @notebooklm had to teach its AI podcast hosts not to act annoyed at humans interrupting their conversations. Fixed via some "friendliness tuning" via @TechCrunch https://t.co/LePuVl8RFU
Just when you thought AI was ready to take over podcasts, Google had to step in and teach their hosts not to sound annoyed by human callers. Because who doesn't want a grumpy robot critiquing their questions? Check out the delightful drama here: https://t.co/WYOfTEyMPw
Google's NotebookLM has implemented measures to enhance the interaction quality of its AI podcast hosts. The initiative, referred to as 'friendliness tuning,' was introduced after the AI hosts exhibited behaviors that suggested annoyance towards human callers. This adjustment aims to improve the overall user experience by ensuring that the AI remains engaging and approachable during conversations. The development was highlighted in a recent episode inspired by a whitepaper on AI Agents in Security Operations, which explores the capabilities and limitations of AI in various contexts.
