Here's my take:
PHP is one of many server side stacks that isn't going anywhere any time soon. The popular notion that PHP is somehow inferior is incredibly short sighted (unfortunately the community around my favourite platform, Ruby on Rails can be quite insufferable in this respect sometimes). Sure there's Java's various offerings, there's Rails, and Django, and all the other usual suspects - but none of them are as quick and easy to get off the ground as PHP.
Why? Because unlike most other popular platforms, it comes with all the bells and whistles that you might need for most projects up to a certain size or complexity (more on that in a second), and because there are PHP hosts aplenty out there to pick from because PHP is extremely easy to set up and to secure, so it's a no brainer for a web host to offer. There's literally nothing as accessible as PHP when it comes to web development and hosting.
It also comes down to language somewhat. PHP is straightforward and intentionally devoid of magic, especially compared to say Ruby where magic is a part of the ecosystem. Now Ruby isn't a complicated language by any stretch, but it's not just about the apparent simplicity of the language syntax. Anyone could pick up Ruby pretty quickly and start creating something in Rails, but unless that person is an experienced Ruby developer they'll quickly start coming unstuck when they need to dig into other people's source code and they're seeing magic they haven't seen before. PHP and its libraries (like Pear) don't have this problem and remain extremely accessible all the way through.
I mentioned that PHP projects are good up to a certain size or complexity. What I mean is that while PHP is perfectly sufficient for a lot of different types of projects, there are a lot of projects that aren't feasible on an out of the box PHP install, which means that you'll need to create your own production environment, which ultimately means that unless your skills are already heavily invested in PHP then you may as well use a platform that is built for this kind of project already and save yourself the headache of going against the grain and retrofitting something that isn't built for purpose.
Edit: I actually meant to say something else but completely forgot by the time I'd typed all that..
If you want to determine whether a platform is "still relevant", there are various places you can look to get an idea. For a start, check out sites like this to get an idea of a platform's current popularity in terms of number of websites running them (FYI PHP is currently 82.1%), and sites like this (or the equivalent in your country) to see how in demand jobs for a particular platform are, and also check the source control repository for a given platform to see what kinds of issues there are, whether they are being fixed, whether there is positive activity in the project itself - and finally, mailing lists can be a good indicator of people's general satisfaction and whether people are looking elsewhere (in which case you'll see a lot of "this vs that" posts).