Actually no, https://twitter.com/summerlinARK/status/1599196885675544576?lang=de. And this is still an underestimate because predicting 10 years in algorithmic advances in the field of AI is silly. And that doesn't even account for distillation, more publicly available datasets and models, multi-LLM systems,... There are so many dimensions in which this train is running, it makes you dizzy thinking abt it and makes regulation look like nothing more then pure cope.