A software startup has always, always been a longshot. This is still true. Also, vanishingly few companies of any kind, including software, ever get a "moat". Small companies survive, when they do, by being fast moving enough to keep making new things.

You certainly should not start a software startup to get rich, as that was never the most likely outcome. You should start one if you have an idea for software that isn't out there, which you think people would find useful, and you want to see what it's like to do a startup (enough to give it several years of your life).

I have, btw, never founded a software startup, though I have worked at some. I got paid in salary, not equity. It is now, always was, and probably always will be a longshot.