There’s been a trend of sorts to use large-scale language models (LLMs), and by extension their outputs, for all kinds of machine learning-driven tasks, including predictive tasks that were already being solved long before language models came into existence.
Source link
