Better results, yes, but that doesn't mean good results. It can only find local optima in a predetermined state space. Training a neural network involves (1) fi

Better results, yes, but that doesn't mean good results. It can only find local optima in a predetermined state space. Training a neural network involves (1) finding the right state space, and (2) choosing a suitable gradient function. If the Correct Solution isn't in the state space, or isn't reachable via gradual improvement, the neural network will never find it. An algorithm that can reason about the meaning of text probably isn't in the state space of GPT. Thanks to the https://en.wikipedia.org/wiki/Universal_approximation_theorem https://en.wikipedia.org/wiki/Universal_approximation_theore..., we can get something that looks pretty close when interpolating, but that doesn't mean it can extrapolate sensibly. (See https://xkcd.com/2048/ https://xkcd.com/2048/, bottom right.) As they say, neural networks "want" to work, but that doesn't mean they can. That's the hard part of machine learning. Your average algorithm will fail obviously, if you've implemented it wrong. A neural network will just not perform as well as you expect it to (a problem that usually goes away if you stir it enough https://xkcd.com/1838/ https://xkcd.com/1838/), without a nice failure that points you at the problem. For example, Evan Miller reckons that there's an off-by-one error in everyone's transformers. https://www.evanmiller.org/attention-is-off-by-one.html https://www.evanmiller.org/attention-is-off-by-one.html If you add enough redundant dimensions, the global optimum of a real-world gradient function seems to become the local optimum (most of the time), so it's often useful to train a larger model than you theoretically need, then produce a smaller model from that.