@everton137 @allaboutberlin Thing is, I would be prepared to bet it's _not_ being trained on their site specifically. I bet it's a generically trained LLM that is live querying Google's cache of their website and "summarising" it (and other related content). In other words, it would bypass the debate about training and copyright. It could be incontrovertibly live copyright infringement.
decoderwheel@hachyderm.io
@decoderwheel@hachyderm.io