The Link: Google News Is Boosting Garbage AI-Generated Articles
The one-line version: Google News promotes “content” regardless of origin, and it’s easier to use generative AI to rewrite other people’s research at scale than to do it oneself, so we’re seeing more — badly written, error-prone, and overall worse — news articles on Google News.
And, I would add: Everywhere. Wherever “content” is an amorphous commodity valued by volume and algorithmic preferences, generative versions of original content is always going to have a competitive advantage, to the short- and long-term detriment of the environment as a source of new information and analysis.
My take: Called it, more than once – a lot of people did. In short: the value of news (or meeting slides or anything else) lies in the original research and analysis embedded in them. Many types of AI can contribute to it, but large language models don’t. Agents maximize utility and original research is slow and expensive, so unless you deliberately reward it, you won’t get it – only piles of bad reproductions of what everybody else already knows.