I've just successfully chatted to ChatGPT about equivalence or at least high similarities between QFT, neural networks and cellular automata (referencing Wofram's work). Does that pattern matching count?
I noticed that if I ask it to tell me how good cryptocurrencies are, it'll do it, and then if I say I disagree and they're wrong, it'll simply switch and agree with me as well. The thing has no care for truth, no opinion of its own, no ability to insist, and just feeds you whatever is statistically close to your own questions.