Microsoft's 'Naughty or Nice' Patent Application 125
theodp writes "Those of you worried about Microsoft's stance on network neutrality won't find much comfort in the software giant's just-published patent application for systems and methods to facilitate self regulation of social networks through trading and gift exchange, which classify users as good or bad and call for network bandwidth to be reduced for those deemed 'less desirable.'"
Re:limitation (Score:2, Insightful)
Re:Will Slashdot be interested in this? (Score:3, Insightful)
It's a website moderation system. (Score:5, Insightful)
Although the patent is questionable, (it sounds similar to the Slashdot Karma System to me) it doesn't sound like something that will be used for net neutrality.
Re:Confused? (Score:3, Insightful)
If it were ever applied to networking, it would most likely be a bandwidth reservation system that gives good uploaders more download bandwidth on a P2P network. That sounds kind of familiar, isn't there a P2P protocol out there that already does this? I can't remember what it's called, something about bits and torrents?
Re:Brilliant! (Score:5, Insightful)
Re:Will it work on Linux/Mac? (Score:4, Insightful)
What happens if a lot of Linux/Mac users give Microsoft a bad rating.
I don't think enough Linux and Mac users could give enough bad rating to MS for it to matter
Re:Slashdot infringes (Score:3, Insightful)
1. a stored sequence of commands for instructing a computing device,
2. such that...
would cover every program ever written. Which, btw, is how every software patent used to start.
Re:Tyranny of the masses (Score:2, Insightful)
Heh. Prior Art? You're read it. (Score:4, Insightful)
*chuckle*
Anyhoo, just what we need -- more technologically-enforced tyranny by majority.
Re:limitation (Score:5, Insightful)
Understanding Survival of the Fittest (Score:3, Insightful)
If you think that evolution and neural nets are doing anything more grandiose, you're in for a rude awakening one day.
The phrase "survival of the fittest" should always cause you to ask "fittest for what?". You should not assume "fit" in this sentence means the kind of "fit" that your doctor (hopefully) proclaims you when you go in for a physical, meaning "fit in all ways". Fittest in the "survival of the fittest" means "capable of surviving whatever hurdle has been put before you today" with no regard as to whether there's any sense of continuity whatsoever to any other hurdle on any other day. Evolution is not cranking out things that are fit for all purposes, it's cranking out things that are fit for the moment, given history only as "how you got there", not proof that you deserve to survive further. The dinosaurs survived hugely longer than man has, and were by all accounts fitter than we'll likely ever be. But then they went away--poof.
Nature favors what's best at the moment, very much like the stock market favors the stockholders of the moment. Nature has no long-term theory of what it is trying to achieve. In a desert ecology, the best design might be the ability to survive without water, but nature can go millions of years designing that model and then if there's a flood one day, nature will favor for survival only those desert creatures that can swim (or maybe that find a cactus to float on), which is not really that different than a corporation buying another just because it likes what's in its bank account and then disassembling the rest for spare parts, even if the part it's disassembling has no long-term value to the population.
Nature always has a myopic view of what it is trying to achieve. It cares about surviving to the next moment, nothing more. Not a lot different than modern corporations caring about surviving to the next quarter, and failing to plan for the long term.
And even neural nets, which you imagine are struggling to be more general, are really hugely dictated in what the will become by what their experience is "growing up". The implicit allegation of the Microsoft patent claim is that they have invented "good parenting, which is the standing "best practice" for training a neural net. Things don't come to be "best practice" without being "prior art".
You might also allege that the claim is equivalent to a perceptron [wikipedia.org], since the notion seems to be that by throttling the bandwidth based on isolated goodness/badness without coordinating activity with other goodness/badness that might operate in a sympathetic way that can generate good results even though it's been pretty well proven that this sort of simplistic system doesn't in fact result in such things.
The problem with patents is that they appear to be a credential. So even though this may be a proven-to-be-bad idea doesn't mean it won't get used. I've often thought of thinking up bad ideas myself and patenting those. They're easier to think up than good ideas, and their being bad doesn't seem to be a barrier to use. If you can get paid (through patent revenue) for other people being stupid, why wouldn't you? You'd think this would retard people moving toward the bad ideas (by making them more expensive) and so implicitly move them toward the good ones, but I fear that the number of bad ideas is so densely packed compared to the good ones that you'd not actually notice any beneficial effect of having lined out only a few of them.
Re:Took a while... (Score:3, Insightful)
And as for the implications of a social networking site downgrading service to nebulously-defined 'bad' users, the effect is quite simple -- you lose those users to your competition. Good-bye clicks, good-bye revenue. Though, of course, 'bad' users could be those that don't demonstrably make money for the site by clicking on ads...