[keeda]

I think what you're looking for is patents. I've said it before, but I think patents are the only protection left for innovative software and "the little guy." It always was, really, but it's blindingly apparent today.

Unfortunately, that would be considered heresy on forums like HN, and people will continue to rail against AI and whatever it's causing and patents, instead of realizing that one is the only available leverage against the other.

[pklausler]

I have a few patents, including one for a novel machine instruction, and I recall the attorney telling me that one cannot patent mathematics, only methods and systems.

[keeda]

That's true but generally that applies to purely abstract mathematics. If the mathematics is truly abstract, no form of IP anywhere would protect it. That has always been (rightfully IMO) the realm of scientific publications.

Otherwise it's straightforward to say that the mathematics is being applied to achieve a practical goal via execution on a computer. (You'll see the term non-transitory computer-readable media" a lot in claims.) You now have a method and system. Now, caselaw frequently changes things, like the "Alice" decision in the US made it much harder to just patent things done "on a computer" but the underlying principle holds.

I'd also guess if your approach makes something faster or cheaper, it should be possible to show it is non-abstract, because resources like time and costs are not abstract quantities.

Standard disclaimer: I'm not a lawyer! I've just worked with patents extensively.