> They have been convinced for years (GPT-2 was released in Feb 2019) that AI is dangerous. A tremendous threat. An apocalyptic threat.

Much longer than that. Sam Altman said it was an extinction threat before he started Open AI in 2015. It's so dangerous that only he, as the best and greatest examples of humanity, was a good choice to create it, and it was even his responsibility to create it to pre-empt some worse person! A fascinating new variation on the White Mans Burden paradigm.