Comment by benchloftbrunch

It's like simple n-gram Markov chain algorithms vs modern LLMs for text