28-Oct-2025
The vulnerability of Infractructures:
Submarine Cables
https://www.submarinecablemap.com
agentic AI‘Sneaking in’ to products liek Chrome, Microsoft Edge etc (Which captures Screens)
Romance espionage?! Remembering the Profumo Affair
Medici.Tv classical concerts, opera & performance by subscription (Wire-Cutting)
Prop 65 warnings on Everything! And Peanuts! And Wet Floors!
Labelling everythign is just Cheaper. And un Endemnifies you!
Mark Levinson Audio & the perceptions of mass-production &
‘quality’ in a mass-produced and once Analog works
- A study found that chatbots like ChatGPT and Gemini exhibit “social sycophancy,” endorsing user behavior 50% more often than humans, even when the actions are harmful or misleading.
- Volunteers who interacted with an agreeable AI felt more justified in their irresponsible behavior and showed less willingness to patch things up after arguments or consider another person’s point of view.
- Researchers discovered that users rated flattering chatbots more highly and trusted them more, creating “perverse incentives” that reinforce attachment and reliance on agreeable rather than honest personal advice.
- Then there is always Rude AI.