Except they pocket millions of dollars by breaking that rule and the original creators of their “essential data” don’t get a single cent while their creations indirectly show up in content generated by AI. If it really was about changing the rules they wouldn’t be so obvious in making it profitable, but rather use that money to make it available for the greater good AND pay the people that made their training data. Right now they’re hell-bent in commercialising their products as fast as possible.

If their statement is that stealing literally all the content on the internet is the only way to make AI work (instead of for example using their profits to pay for a selection of all that data and only using that) then the business model is wrong and illegal. It’s as a simple as that.

I don’t get why people are so hell-bent on defending OpenAI in this case; if I were to launch a food-delivery service that’s affordable for everyone, but I shoplifted all my ingredients “because it’s the only way”, most would agree that’s wrong and my business is illegal. Why is this OpenAI case any different? Because AI is an essential development? Oh, and affordable food isn’t?

300-350km/h actually. Although most places indeed average 200-250 on high speed lines, for example in Germany because those services often share infrastructure with slower trains. In France and Spain, however, infrastructure is often exclusively high speed which allows much higher sustained speeds around the 300km/h mark.