(A similar question got closed as duplicate, but the OP was not exclusively interested in SEO, so here comes an answer about general benefits.)
Including the publication date in the URL can be good URL design.
Usability: descriptive URLs
An URL can give clues about the content it refers to. Giving users such information before they click at the link is a good thing. For example, compare these URLs:
Here the included date can be useful: it’s 2014 now, users interested in current laptops probably won’t find useful information in this post, because the reviewed laptop would be too old.
Of course, this must not be useful for human users only. Data consumers (like search engines) might use this information, too.
I don’t know if any search engines extract it, but search engine behaviour can change often, and there are sooo many of them. I’m only saying that it’s a possibility. An obvious and actual case is highlighting of search terms in the SERPs. See, for example, this screenshot from Google Search where the "2014" is highlighted in the publication date contained in the result URLs:
Usability: browsable URLs
http://blog.example.com/2011/08/21/review-of-my-new-laptop can users allow to browse by removing URL segments from right to left:
http://blog.example.com/2011/08/21/ (all posts published on the same day)
http://blog.example.com/2011/08/ (all posts published in the same month)
http://blog.example.com/2011/ (all posts published in the same year)
(If you don’t want to allow browsing, consider not to use slashes then, and use, for example,
Implementation: name collisions
Some years later, another new laptop. The author is not very creative and names the article, again, "Review of my new laptop". What now? You could add some kind of counter to the URL, like
Not very elegant. In the first example, it looks like the "2" would be part of the article’s title (which is not the case); in the second example, it looks like it would be page 2 of the article; in the last example it breaks with consistency (unless you’d add a
/1/ for all other URLs right from the start).
A common solution for this is to add an ID (like the
49781 for this very question). But why would you want to do that when you could use the date as kind of ID instead? The date has a meaning (and can be useful for users), the ID would be opaque to users (and useless). So it would be
(Under the assumption that no articles with same names are posted on the same day.)
Of course all this is only relevant for websites that have an useful publication date in the first place. Not all pages can/should have a publication date. And even when there are publication dates shown on the page, it doesn’t necessarily make sense to include them in the URL, too.