Point taken. Perhaps my reply could have been more constructive, as well.

The original suggestion to "solve your own problem" is a sound one. He doesn't say that it has to be a problem programmers have. And I doubt he was implying it would absolutely lead to lottery-winning results ;)

I believe that is better advice than finding a problem you can't relate to at all. As Jtsummers also suggests, choose something you will be interested in. Will you want to spend your spare time working on a solution for knitters if you have never touched a piece of yarn in your life?

I would also look at gearing any project towards the B2B market, as opposed to B2C market. I spend much more money with much less hesitancy for tech and software when it helps me or my employer make more or waste less money.

And I hope that most HN commenters can feel free to say "here's what's worked for me" when asked for advice without first offering up a disclaimer, and I suspect most readers can discern a lot for themselves.