You don't have to be unique. The 'next big thing' is not around the corner. The obvious has happened in the software writing market and that is, it has fragmented into specializations. All the main things are being done and have been done for years. It is only the speed and power of PCs that is taking sofware forward and giving commercial advantage to programmers who are first with the fastest, latest iterations of software.

Take a look at Solidworks, it's everything a prototyping engineer could want, but it's $10,000 per licence. Defeating or even competing with Solidworks would take $5 or $6 million. Whereas writing niche market software is extremely lucrative for the programming and support hours that it costs and can be done by a lone programmer.

Also, there are many 'engines' that a programmer can purchase and put his own shell to it. There's lots of 'dead' software out there too where the author has placed it in the public domain, sometimes with source code and if you purchase that, modify it to your own purpose you can sell that if it's changed in a sufficiently dramatic way. There's also lots of free source code out there that can help with programming whatever you want.

Some programmers are stuck with having to give away every version until they think they have got something good and then start selling it at version 6 for $25. That's an ok way to go but not the best way if you want a faster return on your programming effort.