Fixing IP laws, part 5 - Policing Patents

Series links:

Patents Are Being Misused

Let's start off with that simple statement - patents are being misused.

There are some good things about the patent system. The patent term is short, especially when compared with the ever-extended lengths for copyright.
OK, when I said some things, I meant one thing...

I can't think of anything else. I really can't!

It's Not All Patents

In some areas, patents are generally fine. For example, in the arena of Chemicals, patents have struck a reasonable balance between protecting an innovation so that the inventor can recoup costs and make a profit, and allowing future use by others.

But in computing, patents have been nothing by a blight. And that is where I will focus my efforts.

Software and Business Method Patents Must Die

Software and "business method" patents do no good. They are, in my opinion, a perversion of the idea of a patent.

A patent is supposed to strike a balance. In exchange for telling everyone how something is made, the inventor gets a limited monopoly on making it.
Prior to patents, we had Guilds - and they controlled both the production AND the knowledge of how to produce. They distorted the market, and if a Master died before teaching his apprentices, then his knowledge was lost until someone else figured out his "tricks".

But with patents, all society knows how to reproduce your invention - they just can't do it legally.

The alternative is no patent, keeping it a trade secret, and hoping nobody else figures it out - because you'll have no protection.

So, if you invent a new type of gear linkage, or a new way to make nylon, you patent that component or process, and get to control its use for 14 years or so.

But software and business methods patents aren't attempting to control physical objects or processes with physical objects. They're attempting to exert control over the ideas themselves.

Go look at a random software patent - you'll rarely see source code. Instead you see diagrams with "black boxes" which describe the process.

And what use is that to society? How are we supposed to reproduce these mysterious black boxes? Is there a reference implementation that you can show, or are you just patenting a wishful thought?

This is the problem with such ephemeral patents, and it is exactly why software and business method patents should not be allowed.

Fixing Patents

I'm not a patent specialist, and I don't want to risk reducing the value of patents in industries that deal with physical objects and processes. But I do have some suggestions.

A patent is supposed to be "non-obvious to someone with expertise in the art". In other words, any other engineer or chemist should be able to look at it and say "Neat, I wish I'd thought of that."
The sum of human knowledge is now so vast that the examiners who try to check the patent is indeed innovative don't have a chance.

So maybe we should consider changing the law so that any one claim in a patent being found to be invalid will invalidate the patent as a whole?

Sure, folks will submit more patents. But these days, some patents have 30 claims or more. Let's try to take that down to a smaller number of higher quality claims, and dissuade folks from using a blunderbuss to write their patents when they should be using a rifle...

Summing Up

Software patents shouldn't exist. Business method patents shouldn't exist. And patents in general should be a higher quality with fewer claims per patent.