People spend resources trying to discover things because they want to use them. If it turns out that when you go to use your invention, a patent troll jumps out from under the bridge to shake you down, people will be less inclined to spend resources trying to invent new things to use and instead just stick to the status quo.

This makes sense when the price for discovery itself is very high compared to putting it into use. In many fields (especially IT/software/algorithms), the discovery/invention often isn't the main part of work (especially not when you look at what are the actual claims of many inventions in those fields), but creating a good implementation and/or improving on it.

There's a reason that software and algorithms weren't deemed patentable for a long time (and still aren't in certain jurisdictions - at least on paper).