Been a lot of talk on the internet about the Chinese are getting ahead of us at inventing new products and new weapons. We here in the United States were the world’s champions of inventiveness for many years. Even so, there are a few things we could do to improve the situation.
First of all we could clamp down on patent trolls. Anyone who introduces something new will be sued for patent infringement just as soon as something new looks like it will make money. The trolls, pretty much all lawyers, hold vaguely written patents with claims as broad as all outdoors, pop up out of the woodwork and demand money.
We need to reform the patent office. No patent should be granted unless it describes a device in enough detail as to permit someone to make said device, and that the device works after it is built. This means the patent application shall contain a drawing with dimensions, a test procedure so you can see if it is working, a parts list, an electrical schematic, and a parts list. Patents of devices containing computers or micro processors shall contain a listing of the computer program. Patents that lack this detail will not be granted.
All patent applications shall be publicized so that interested parties, often competitors, have time to review the patent application and if it conflicts with patents they already hold, or is ordinary prior art well known in the trade, they can make objections. The patent office shall hold a meeting on each patent application where interested parties can voice their objections to granting the patent.
Interested parties may sue the patent office to repeal old patents which do not conform to these new guidelines.
Much research is conducted by universities, using grant money. A distressing number of papers published by these scientists fail to duplicate. In other words when other people attempt to duplicate the reported results they cannot do it. This is so common that I encountered it myself. I had searched the literature, found a paper describing a method that we needed. I spent a week coding it up and debugging it. It worked, only the performance was only half what the paper had claimed for it. I finally telephoned the author and asked him what I was doing wrong. The author, somewhat sheepishly, admitted that he had left out a factor in his computation of efficiency, and the method was only half as effective as claimed in the paper.
This sort of thing wastes a lot of time and money. Grantees ought to keep track of research they funded that fails to repeat. The scientists who publish fails-to-repeat papers should become ineligible for any more grants.
Doing just these two things would increase the number of new things invented and developed, a lot.