What we really need is another paradigm shift in AI. Current transformer models are great, but they’re clearly not quite what we’re looking for.

That said, an AI breakthrough on that scale is about as likely as an energy breakthrough, so I guess there isn’t much practical difference, really.