[Farer]

It would be great if it could be applied universally, but it seems that accommodating all situations in the real world won’t be easy.

In the end, I feel like it might have to transition into the realm of inference, much like AI that mimics human reasoning.

[malux85]

This has been done, many times before, it turns out the neural network just learns a crapper version of A*, and of course, any domain knowledge from whatever environment it's in, there was a post on this very thing last year on hacker news.