3 ms·

I spent a week vacation with 2 friends who work at the patent office, so w e had time to discuss over beers. They don't believe there is a problem. They rely

I spent a week vacation with 2 friends who work at the patent office, so w e had time to discuss over beers. They don't believe there is a problem. They rely so heavily on prior art, and have trouble with the very concept that something was too obvious for the first 20 programmers to want to patent it, so prior art is very incomplete. I show them 1 click, they tell me that if it was that obvious someone would have patented it before. They refuse to consider the "skilled in the art" obviousness test because they say it's too easy to believe something is obvious after it's been explained to me.

abeppu 15y ago
I can believe it's too easy to believe that something non-obvious is obvious once you've understood it. But that doesn't mean that some things aren't legitimately obvious. It seems like we ought to take an empirical approach to distinguishing which are which -- take skilled engineers either in groups or singly who haven't heard of a given solution, pose them the problem that the solution is meant to solve, let them brainstorm or think for some fixed amount of time, and see if anything they come up with largely reproduces the solution in mind. If fewer than k engineers are needed before someone thinks of it, then it's obvious. If more than l > k engineers are needed before someone thinks of it, then it's not obvious. If somewhere between k and l engineers are needed, then maybe you still get to argue about it. On a related note, I think I have a new strategy for generating interview questions.
brlewis 15y ago
They refuse to consider the "skilled in the art" obviousness test The law includes an obviousness test for a reason. If Congress intended prior patent applications to be the sole test, they would have said so. The patent office is ignoring the law to the detriment of all of us.

abeppu 15y ago

I can believe it's too easy to believe that something non-obvious is obvious once you've understood it. But that doesn't mean that some things aren't legitimately obvious. It seems like we ought to take an empirical approach to distinguishing which are which -- take skilled engineers either in groups or singly who haven't heard of a given solution, pose them the problem that the solution is meant to solve, let them brainstorm or think for some fixed amount of time, and see if anything they come up with largely reproduces the solution in mind. If fewer than k engineers are needed before someone thinks of it, then it's obvious. If more than l > k engineers are needed before someone thinks of it, then it's not obvious. If somewhere between k and l engineers are needed, then maybe you still get to argue about it. On a related note, I think I have a new strategy for generating interview questions.

brlewis 15y ago

They refuse to consider the "skilled in the art" obviousness test The law includes an obviousness test for a reason. If Congress intended prior patent applications to be the sole test, they would have said so. The patent office is ignoring the law to the detriment of all of us.