Conventional wisdom these days is that it's impossible, and moreover, you don't even have a chance of success unless you're backed by a large corporation that can afford to blow several million on development.<p>Conventional wisdom could always be wrong - there's always a chance that the environment could shift and up-end everything we know. But the last programming languages designed by an individual to gain widespread acceptance were PHP and Ruby in 1995, Python in 1991, and Perl in 1987, and in most cases widespread success came only 15 years after invention. Other than that, everything's been corporate: Swift (Apple), Rust (Mozilla), Go & Dart (Google), Javascript (Netscape), Java (Sun), Objective-C (Sun and NeXT), C# & Visual Basic (Microsoft), C & C++ (Bell Labs). You could argue that Scala (Martin Odersky, 2004) and Clojure (Rich Hickey, 2007) might count, but I wouldn't call them mainstream in the sense of "You can have a safe career knowing only that language."<p>I think your best bet for building a programming language for a living is to become independently wealthy by founding a startup, and then write the programming language afterwards. Or alternatively, build a critical system for a large big company so that you're super valuable and have a track record, and then threaten to leave if you can't do your pet programming language project.