The author is right about many of the patent system's problems. But none of the problems mentioned have much to do with Theranos specifically, apart from its c

The author is right about many of the patent system's problems. But none of the problems mentioned have much to do with Theranos specifically, apart from its convenient availability as narrative material. Theranos was a biotech company. Biotech companies as a rule are patent-focused, patent-buoyed, and patent-dependent. The high R&D costs of biotech and pharma might be the best fit for the patent system we have. Software might be the worst. Under US law, inventions must be new, useful, but not obvious to be patented. The article rightly points out that the Patent and Trademark Office does not spend lots of time or money assessing usefulness beyond a gut plausibility check. We can assume they approve patents on useless inventions every year. But what is the use of a patent on a useless invention? You can still license others to practice your useless invention. It may take some time for would-be customers to realize it's bunk. But in the end, nobody will rush to use your idea without a license, or to buy a licensed embodiment from you. The USPTO is not a broad-remit anti-fraud organization. If that's anybody, it's SEC. It's not USPTO's purpose to vet new patent applications for commercial viability. That kind of due diligence is customers', investors', and potential employees' job, on the front end, and perhaps FDA's and FTC's, as backstops.