How many of those are there? I give talks pretty frequently where I start by asking people in the audience what tools they regularly use for technical computing, and I'm legitimately surprised the few times anyone says they use Octave for things beyond classroom-level work.
No offense intended to its developers, but Octave as a project seems like misdirected effort to me. Matlab the language isn't the most useful thing to try to emulate, Matlab the environment and tool set for getting non-software engineering jobs done is why people use it. For example, an increasing portion of the people who still use Matlab do so largely because of Simulink, for which there isn't a comparably effective open source alternative yet (you can sometimes use Modelica but there's less overlap in users than you'd think).
As much as I detest the MATLAB licensing reaming (and sometimes the inelegance of the language itself), there is very little out there approaching the sheer depth of the associated toolboxes and Simulink models.
- Octave won't die. It won't die even if you ditch it altogether. A lot of companies depend on it because licenses are free and that makes it easy to do massive parallel computations by launching thousands of instances. Matlab doesn't let you do that. As long as it stays GPL, the employees within those companies and academic institutions will continue to contribute to it. (Python is competition, but there are still plenty of Octave supporters.)
- You might consider working for a company that is built around Octave. They would likely gladly welcome you as well as support you in spending a percentage of your time to maintain it, since their business depends on it.
- Build something that people will pay for. Like a decent GUI for Octave that does everything Matlab's GUI does. Or a few awesome closed-source toolkits. That alone might support your personal income needs.