4 ms·

>had some random idea I had 20000 times in my life already, Aren't patent systems already supposed to reject "inventions" that are common sense? Perhaps the pr

>had some random idea I had 20000 times in my life already, Aren't patent systems already supposed to reject "inventions" that are common sense? Perhaps the problem is not the system itself but rather the humans who are approving these.

acdha 3y ago
That’s definitely part of it: in the 90s you saw a ton of “on the internet” patents which looked a lot like the examiners not having had enough experience to say that something was either an obvious adaptation of an existing concept to the web or that the two systems had almost nothing in common other than using a computer (I remember a patent troll hitting a customer with a 1980s patent for cash registers connected to a minicomputer using a phone line, claiming it covered their web store). A younger generation of patent clerks hired with more experience seems to have helped there. The other problem is funding: look at the patent examiner listings right now and think about how many people they’re going to be able to get with the federal pay scale having been prevented from keeping pace with the market for years: https://www.usajobs.gov/Search/Results?l=&l=&a=CM56&hp=public&k=patent%20examiner&p=1 https://www.usajobs.gov/Search/Results?l=&l=&a=CM56&hp=publi...
lostdog 3y ago
The parent system will accept pretty much anything if your lawyers ask them enough times. I've seen algorithms from the 80s patented today, with no real changes. The patent examiners have no idea what's novel or common sense, and just accept everything.
brlewis 3y ago
It is not reasonable to expect humans to assess the novelty or obviousness of software "inventions". Too many of them can be created too fast for any conceivable patent office to handle. The only solution is for Congress to write a law saying algorithms can't be patented. SCOTUS tentatively said it already, but nobody listens to them.

acdha 3y ago

That’s definitely part of it: in the 90s you saw a ton of “on the internet” patents which looked a lot like the examiners not having had enough experience to say that something was either an obvious adaptation of an existing concept to the web or that the two systems had almost nothing in common other than using a computer (I remember a patent troll hitting a customer with a 1980s patent for cash registers connected to a minicomputer using a phone line, claiming it covered their web store). A younger generation of patent clerks hired with more experience seems to have helped there. The other problem is funding: look at the patent examiner listings right now and think about how many people they’re going to be able to get with the federal pay scale having been prevented from keeping pace with the market for years: https://www.usajobs.gov/Search/Results?l=&l=&a=CM56&hp=public&k=patent%20examiner&p=1 https://www.usajobs.gov/Search/Results?l=&l=&a=CM56&hp=publi...

lostdog 3y ago

The parent system will accept pretty much anything if your lawyers ask them enough times. I've seen algorithms from the 80s patented today, with no real changes. The patent examiners have no idea what's novel or common sense, and just accept everything.

brlewis 3y ago

It is not reasonable to expect humans to assess the novelty or obviousness of software "inventions". Too many of them can be created too fast for any conceivable patent office to handle. The only solution is for Congress to write a law saying algorithms can't be patented. SCOTUS tentatively said it already, but nobody listens to them.