disc: former Amazon employee
disc: former Amazon employee
The idea is that a patented invention is novel enough that it's unlikely someone else would independently invent it, so the patent encourages the inventor to share all the details of the invention (and in exchange gives them a monopoly on it). Patents are specifically _not_ for good ideas that aren't non-obvious inventions. You can't patent the idea of "1 click buying" or "a time machine", only an actual invention/implementation that does those things.
I agree amazon's 1-click experience is a really good idea. It's also well designed and implemented. The question though is whether some other developer could, given the idea of "1 click buy", create this "invention".
I think it's unquestionable that, given the problem statement, many other developers of the time could have created the same thing.
In fact, a bunch of other developers have already created 1-click buy buttons, and I assure you the vast majority did not read amazon's patent to figure out how to write that code.
The fact that people can build this invention without looking at amazon's patent, to me, proves it should not be patentable. The general public gained nothing from the patent.
To get that swap you need public disclosure in some industries, and to get public disclosure you need incentives. For widely available products that disclosure may consist of the actual object without any patents. But, for say oil refineries people don’t just get to walk around and see what’s what.
Having said that the parent system could be changed in a vast number of ways with many tradeoffs. Just consider the wider implications beyond just software or mass produced items.
If the patent was never invented and filed, would a child be able to come up with and communicate the idea patented?
If one click to buy didn’t exist today, how many children would be able to simply state the obvious solution of one click purchasing?
You also are talking about the idea, not the invention. The idea was not patented, the technical invention (effectively flow-charts for how to have a button click trigger an auth cookie check, have databases retrieve credit card numbers, etc), which is not obvious to a child.
But, I would argue, the details of the invention would be clear enough to someone skilled in the art, another competent programmer.
That seems like a strawperson. The parent said that just given the very rough idea or (I would add) even just the term "1 click buy", somebody could have created a comparable experience without any reference to the patent. That's not "give someone enough constraints". That's "let somebody overhear a customer mention the feature in passing".
Suggesting that Amazon wouldn't have invented 1-click buy without patents seems incompatible with good business practices. All my conversations with other developers mentioned a single reason for patenting "random" features at this point: As a way to defend against other people's patents. The only reason _not_ to create a novel feature was that it may be too close to something patented and a lucky lawyer may have argued that it was too close. If anything, software patents actively discourage innovation.
I've never heard of anybody implementing a feature by using a patent. I've never heard of somebody developing a novel feature to get a patent. Which means as "patents", they are useless.
Yet, overhear the idea in passing and it’s implementation is obvious. Actual secure crypto is much harder than a simple implementation though. RSA is interesting because an equivalent algorithm was independently discovered in 1973 at GCHQ the British signals intelligence agency, and only declassified in 1997.
So in effect it’s disclosure not invention that shaped public crypto. I am not saying patents are a great fit for software, but I am saying disclosure in one form or another is.
In the one-click case, it's like getting a patent on the idea of conversion of heat to mechanical motion, then demanding payment from James Watt for his steam engine.
IIUC, the standard is regrettably non-obvious to someone of ordinary skill in the art. "Sure it obvious to her, but she's an expert!" can be a valid defense of patent non-obviousness. :/
> A skilled person is typically defined as an expert in a relevant field who has average knowledge and ability but is not exceptional, outstanding, or brilliant. Such a person must be neither over- nor under-qualified and is deemed to be uninventive.[1]
> This raises a difficult question of how to define an appropriate benchmark for AI-assisted inventions and whether/how the current standard should incorporate AI.
[1] https://ijssppnet.com/file/2.%20%20%20%20Dr%20Olga%20Gurgula...
I totally disagree with that. There's even apocryphal story that when 1-click purchase was implemented, developer actually couldn't resist of adding "are you sure" message box, because just click and buy is so against typical patterns.
I can give another example - socket wrench. Ratcheting wrench was patented on 1863. It took 110 years to invent quick change socket for it, although if you look at it you will say - well, everybody could create that invention.
In my opinion, it’s a stupid idea and I think anyone could have come up with it. No one did it because it’s an anti-pattern.
1 click buying probably fails that. I'm on the fence about "lossy compression of audio files using wavelets, based on a psycho-acoustic model of hearing range" (aka mp3)
I think it was obvious at the time, and cannot fathom why other stores didn’t try it earlier except perhaps fraud fears.
Separately, Amazon is often no longer one-click, which will bite you when you think you ordered but were instead on a supplemental question page you clicked back from without reading. Super annoying when the product doesn’t show up 2 days later.