That may be true in other fields, but that argument makes some assumptions about invention and production that don't hold true for software. First, the notion

That may be true in other fields, but that argument makes some assumptions about invention and production that don't hold true for software. First, the notion that production is so expensive that a lone developer cannot realize their idea. This isn't true. Any developer can build and market a product single-handedly, in the majority of niches of the software market. Secondly, the idea that a patentable idea is born first, and then a product is built based on it. That happens now and then, but it's rare (e.g. RSA patent). The common case is that a developer sets out to build a better mousetrap, delivers a piece of working software, and only afterwards realizes they've had a patentable idea. By that time they also have a marketable product.