I get why that's the default behavior but "adds nothing of substance" is a huge generalization that fails in this case.

An article with zero details about the proof could easily be titled "New Orleans Teenagers Found a New Proof of the Pythagorean Theorem" but couldn't accurately be titled "Here’s How Two New Orleans Teenagers Found a New Proof of the Pythagorean Theorem." Whereas here it says it'll have more details and it does. (although TBH this is less a "here's how they found" and more a "here's what they found", if I'm being extra pendantic).

The shorter title is less descriptive in this case.

This is getting off topic, but IMHO the HN default behavior here is reasonable.

There's certainly situations where removing the leading "Here's How" makes the title worse, but I think those instances are rare and in general this rule leads to better titles much more often than worse titles. Manual human review would of course be better, but dang only runs in O(n) time. Basically, it's not perfect but I think it does much more good than harm.

It also doesn’t tell you “how” either. Almost never does an article tell you “how” or “why” despite it being the first word.

> (although TBH this is less a "here's how they found" and more a "here's what they found", if I'm being extra pendantic)

Yeah, I guess I'm a bit extra pedantic when it comes to the titles, because I agree the most with this, the article doesn't seem to actually go into how they found it out, meaning the original title was misleading after all.

just taking out the "here's" and leaving "How New Orleans Teenagers Found a New Proof of the Pythagorean" would have preserved the meaning and also made it seem less clickbaity.

No. The short version titles an assertion, the longer version titles an exposition.

HN automatically strips things like this on submission but you can edit the submission title after posting to put it back in when it's appropriate, as here.