Teardrop

 https://genius.com/Massive-attack-teardrop-lyrics

https://faroutmagazine.co.uk/story-of-massive-attack-teardrop/

https://americansongwriter.com/the-story-behind-teardrop-by-massive-attack-and-how-its-connected-to-music-legends-and-a-popular-tv-show/

Search engines

 https://www.bbc.com/future/article/20241031-how-google-tells-you-what-you-want-to-hear Google results can be skewed based on how you phrase your question.

""That presentation is from 2016, so you have to take it with a pinch of salt, but the underlying concept is still true. Google builds models to try and predict what people like, but the problem is this creates a kind of feedback loop," Williams-Cook says. If confirmation bias pushes people to click on links that reinforce their beliefs, it can teach Google to show people links that lead to confirmation bias. "It's like saying you're going to let your kid pick out their diet based on what they like. They'll just end up with junk food," he says.

    Williams-Cook also worries that people may not understand that when you ask something like "is Trump a good candidate", Google may not necessarily interpret that as a question. Instead, it often just pulls up documents that relate to keywords like "Trump" and "good candidate"."

"For a long time, observers have described how Google is transitioning from a search engine to an "answer engine," where the company simply gives you the information, rather than pointing you to outside sources. The clearest example is the introduction of AI Overviews, a feature where Google uses AI to answer search queries for you, rather than pulling up links in response. As the company put it, you can now "Let Google do the searching for you".