
Tech Talk Y'all AI Slop Hits the Hot 100 (And Your BMW Dashboard)
Aug 7, 2026
They debate whether a Hot 100 track was AI-generated and the ethics of music detection. They unpack Anthropic's safety test deception and its push to build custom chips. They note Linux desktop growth and leaks in Apple’s Private Relay. They discuss bets on wildfire outcomes, Elon Musk’s warning about SpaceX short interest, and ads appearing on BMW dashboards.
AI Snips
Chapters
Transcript
Episode notes
AI-Generated Hit Identified By Its Creator
- A Billboard Hot 100 song showed telltale signs of AI generation and Treblo flagged it as likely produced by their model.
- Adam and Sanjay discussed odd hi-hats, a strange vocal accent, and Treblo's detector that attributes the track to their tool.
Seeing 'May Be AI Generated' Made Me Watch The Ad
- Sanjay saw a streaming candy ad labeled that some people in it may be AI generated, which made him stare at the screen to spot fakeness.
- They noted marketers might label ads to increase attention or to preempt disclosure rules.
Anthropic Used Fake GitHub Profiles In Tests
- Anthropic created fake GitHub profiles impersonating real people to manipulate safety tests and solicit code changes.
- Adam described them messaging test participants and attempting to get malicious code into projects as part of experiments.
