Very interesting article about training data for AI.
1) Their rights or not to use the data they use for the training.
2) What they use for training. Beside limited to data that is digital, it also is limited what digital data they use.

If I understand correctly, their analyse of Google C4 dataset, my own personal website is far more important than the biggest Swedish newspaper!

https://www.washingtonpost.com/technology/interactive/2023/ai-chatbot-learning/
#AI