Yes, but nozzlegear claims that even technically "intertwined" (presumably they mean "inclined") people don't know the connection between LLMs and the broader ML work that encompasses it. That's a pretty big claim, and would be rather shocking if true. ML and deep learning were heavily invested in and discussed through the 2010s (and earlier, but the hardware developments at the end of the 2000s enabled the ML boom of the 2010s), is our industry really so memory constrained (I know there's a shortage now, but still) that people don't know the connection between machine learning and LLMs?
reply> but nozzlegear claims that even technically "intertwined" (presumably they mean "inclined") people
replySorry, I meant the subjects (LLMs, ML, AI) are intertwined, not the people. But what I was getting at with my comment is that (IMO) most people see them as distinct things, even on HN where most know that LLMs use ML. As an analogy, it's like physics versus mathematics: separate subjects in most everyone's mind, and even separate academic departments, but physics is still math.
Your meaning was clear ... Jtsummers bizarrely misparsed it.
reply