[mattigames]

That's the problem, in technical links popularity (as in number of visits) is not linked to correctness, that's one of the reasons StackOverflow is so useful, because the top-voted answer is like 99% of the times the most correct (useful) one.

[ysleepy]

Except when that answer is 5 years old and is as wrong as the ExpertsExchange post from 2001.

[ceejayoz]

Unlike EE, though, any user can propose an edit to that now-incorrect answer (or just edit it outright, if they have 2k rep).

A nice big "this answer is outdated, see [other answer]" note does wonders.

[JeremyBanks]

Unfortunately, such edits are rarely accepted. The intent was for the site to be more wiki-like, but these days "respect the author's intent" is the prevailing wisdom. The cases where someone edits a top answer with a significant correction or update are very few, and require overwhelming community support or they'll be reverted.

This is particularly distressing in the case of security vulnerabilities. Vulnerable code being copied from Stack Overflow into consumer applications isn't hypothetical, it's been empirically demonstrated (https://www.aisec.fraunhofer.de/en/stackoverflow.html), yet the community and company refuse to accept responsibility and address the problem. (Of course the developer creating the application is more responsible, but Stack Overflow is in a position where they could help protect innocent end users and the common good.)

Pinning of accepted answers to the top of the list was a good idea when the site was young. These days, it prevents updated better answers from taking the place of ten-year-old misinformation.