You can't call any piece of evidence that happens to not support your theory as being an outlier. And even then the OpenAI saga is probably the most important development in tech in the last 10 years (conservatively).
Perhaps, in terms of how much it's affecting Silicon Valley; though I'd say iPhone adoption, the tail-end of Flash, the death of ActiveX, the destruction of libraries, and the growth of Amazon were all more impactful, there.
Probably not, in terms of how much 2040s tech will be based on this stuff. Language models are good for machine translation, and real-time image transcription, but everything else I've seen them do has better solutions (which have been around for decades in many cases, but don't have much funding).
