I came across this post a few days ago:
The Great Programming Language Continuum; or why C++ is dying
As someone who programs in C++ frequently and who is fairly connected to the C++ community I think that this author really has no idea what he is talking about in this case.
I first think that the author's list of languages from lower to higher levels is quite flawed. Following the list and his logic of how users strive for higher levels everyone would write everything in lisp if they could... This seems to mesh strangely with how language adoption looks among new developers based on statistics.
I also think that the point, which others also seem to make every now and then "Hardware advances will continue to make its efficiency redundant" is quite a weird statement. It is true that computers get faster, but as that happens our demands of what we want them to do also go up. For random desktop applications such as an e-mail client it may not matter that much, but for the next Crysis you will definitely want to squeeze every percent of power even out of the best computer in 2015.
It is also the case that C++ is taught extensively in schools everywhere, and engineers will reach for the first tool they think of when solving problems. Not to mention that from personal experience well-trained engineers writing good C++ can be almost as quick as engineers coding in Python and write code that has no more bugs.
There are more and more massive open source projects in C that adopt C++ for parts of their project, even including long-running massive projects like GCC. This along with having very popular frameworks like Qt with tons of contributors definitely signals interest in the open source space is still strong.
Interest in business is also definitely still there. In areas where high performance software is needed like for example game development C++ has pretty much completely taken over from C due to the flexibility and readability of code it allows for in large codebases.
What about the data then? Doesn't that contradict my statements?
First of all, C++ has gone up about 1% according to the chart... so I'm not sure what the data is supposed to prove. I do think C++ has lost some of it's "flair" meaning it shows up less in searches of the internet, which is why it hasn't grown more. I do think C++ is not likely to go to the top of any popularity contest, but I don't think it is going away anytime soon. I even think we're likely to see many other languages above C++ in the TIOBE ranking go away first.
I think the author is a bit blinded by his allegiance to the languages he uses, which I probably am as well to some extent, but I think that in this case I hold more facts than he does.