This excellent blog article describes a technique developed by Judea Pearl decades ago to do exactly this. Would be interested to understand how this is different/better.
'the harm to consumers that Microsoft's decision could create.'" The only harm is to these business' pocketbooks.. For once I'm on MS side in this matter...
Then you are failing to think this through. The only effect of this will be to give advertisers an excuse to ignore "do not track". How does undermining "do not track" help anyone?
For once I agree with Microsoft that WE DO NOT BENEFIT FROM TRACKING.
Are you stupid, or can you just not read? Microsoft is UNDERMINING "do not track" by turning it on by default, because that only gives advertisers an excuse to ignore it. "Do not track" is entirely voluntary on the part of advertisers.
Because it renders "Do Not Track" useless. Apache is already ignoring Do Not Track if it detects that you are using IE10. It's a boneheaded move on Microsoft's part.
The Tahrir Project is trying to create an anonymous microblogging platform, similar to Twitter or Facebook. Google was sponsoring development on it over the summer so with any luck it won't prove to be vaporware like Diaspora.
The plan is that people will be able to define lots of functions like this, along with much more complicated ones, and then share them. The best of them will become part of the default vocabulary.
Please sign up for the mailing list if you'd like to keep up with developments (or, if you can code Java, perhaps you could help?!)
Yeah, it kinda is. Did you ask that when Slashdot opened their codebase many years ago? How about when Reddit did it? What about Google with their various open source projects?
You should be glad that people open source things.
Soulver was actually what inspired LastCalc, but I wanted to bring it to the web, and make it programmable.
OpalCalc looks neat, unlike Soulver it supports functions, and I'm sure it has a few features that LastCalc currently lacks.
However LastCalc has a few features that OpalCalc lacks too, such as support for higher-level datastructures like lists and maps, pattern matching (like Haskell), and the ability to pull data from the web to use in calculations.
So I'm not sure that I would describe OpalCalc as "LastCalc on steroids" by any stretch.
That said, Tesla is a US company, and Top Gear Is produced by BBC in the UK. So I'm not sure there is much hope for this lawsuit to accomplish anything anyways.
I'm guessing you haven't heard much about British libel laws.
Why are they coming up with their own operating system and app ecosystem, is this really the core competency of a car company?
Why aren't they using Android, which already has text to voice, voice to text, GPS navigation, and almost everything else you might need in a car?
Slashdot no-longer looks like a car crash by reverting to essentially the level of site complexity they had in 2006 (or, at least, disguising the "improvements" more effectively).
Will be interesting to see if this stems or reverses the exodus of readers/. has experienced over the past half-decade.
Rather it is an auto-immune disorder, gluten causes the sufferer's own immune system to attack their small intestine. Aside from the immediate pain (a stomach ache for a day or two after eating even a tiny amount of gluten), it can result in deficiencies in various necessary substances, and can lead to an increased danger of cancer.
My wife had stomach aches most of her life, she had grown accustomed to them, thinking they were normal. A few years back (she was 28), on someone's suggestion she got tested for Celiac (first a blood test, then a biopsy of her small intestine). She was positive.
She has been avoiding gluten ever since, she can't even have a single crumb of bread without getting sick now.
Most people that have Celiac are never diagnosed, and suffer a life of pain and misery as a result, in addition to a shortened lifespan. If you get a lot of stomach aches for reasons you can't determine, you owe it to yourself to get a blood test for Celiac.
Re:Taking care of people is not wrong
on
Health Care Reform
·
· Score: 1
It's about forcing you to do these things at gunpoint (and yes, a gunpoint is somewhere in your future if you stop paying your taxes) by raising taxes (by 3.8%) and by forcing you to buy health insurance when you don't want to do so.
And yet I'd bet you'd be the first to expect to get treated in an emergency room if you didn't have health insurance and something bad happened. The reality is that nobody wants to live in a country where people are allowed to die on the streets because they are poor. If you accept that as a premise, then at some point it is necessary for us all to accept some kind of mandate to participate in the health care system. The alternative is the current situation, the worst of all worlds, where emergency rooms end up being the safety net for those without insurance, and we all pay through the nose for it.
One reason this kind of problem occurs is that many collaborative filtering algorithms are measured based on "root mean squared error", basically the square root of the mean of the differences between what was predicted and what the user actually did.
The problem with this metric? It doesn't account for a variety of important things, one of which is that most users value diversity. Another is that in most recommendation systems, what is important is the relative relevance of recommendations to each-other, whereas RMSE is an absolute measure of effectiveness. And a really tricky one is that the recommendation algorithm itself can impact user behavior. For example, the user may raise their standards if the algorithm does a better job.
The unfortunate answer is that the only rock-solid way to measure the effectiveness of recommendation algorithms is to test them with real users, perhaps splitting the user population between different algoritms, and seeing which does best.
I'm pretty familiar with this issue as my day job is building a behavioral ad targeting engine. We learned a long time ago that while RMSE has its uses, there is often limited correlation between an algorithm's ability to predict user behavior retrospectively (which ads they will click on and what products they will buy), and how much additional revenue the algorithm will generate in practice.
Our solution is to use RMSE as a first-blush indication of how good an algorithm is. Secondly, we take the top, say, 10% of ads with the best predictions, and see what the actual click or conversion rate is within this 10%. This requires a higher volume of data, but yields results that are closer to what we find in reality. Lastly, the algorithm then has to prove itself in the wild on a small subset of traffic. Only then can we really know if any algorithm is an improvement on any other.
It seems that Wikileaks should operate over Freenet. Leaks could be submitted anonymously that way, and also distributed anonymously. The advantage would be that it would be entirely decentralized, so there would be no organization vulnerable to legal action.
Freenet has been slow and hard to use in the past, but its improved quite a bit. It is the obvious platform for something like Wikileaks. Of course, there is nothing to prevent people from mirroring content on the web (since installing Freenet, like any piece of software, is a hassle). But at least there will be an unimpeachable backup of all data on Freenet.
Its unsolvable to do it in advance, but quite possible to do it while observing the running code (a bit like how a filesystem optimizes the locations of data on disk).
Slashdot Mirror
This excellent blog article describes a technique developed by Judea Pearl decades ago to do exactly this. Would be interested to understand how this is different/better.
Then you are failing to think this through. The only effect of this will be to give advertisers an excuse to ignore "do not track". How does undermining "do not track" help anyone?
Are you stupid, or can you just not read? Microsoft is UNDERMINING "do not track" by turning it on by default, because that only gives advertisers an excuse to ignore it. "Do not track" is entirely voluntary on the part of advertisers.
Because it renders "Do Not Track" useless. Apache is already ignoring Do Not Track if it detects that you are using IE10. It's a boneheaded move on Microsoft's part.
The Tahrir Project is trying to create an anonymous microblogging platform, similar to Twitter or Facebook. Google was sponsoring development on it over the summer so with any luck it won't prove to be vaporware like Diaspora.
You can just type:
X nm = (X*10) angstroms
The plan is that people will be able to define lots of functions like this, along with much more complicated ones, and then share them. The best of them will become part of the default vocabulary.
Please sign up for the mailing list if you'd like to keep up with developments (or, if you can code Java, perhaps you could help?!)
Or somehow silence grammar nazis...
Yeah, it kinda is. Did you ask that when Slashdot opened their codebase many years ago? How about when Reddit did it? What about Google with their various open source projects?
You should be glad that people open source things.
Nah, just wanton irresponsibility :-)
Soulver was actually what inspired LastCalc, but I wanted to bring it to the web, and make it programmable.
OpalCalc looks neat, unlike Soulver it supports functions, and I'm sure it has a few features that LastCalc currently lacks.
However LastCalc has a few features that OpalCalc lacks too, such as support for higher-level datastructures like lists and maps, pattern matching (like Haskell), and the ability to pull data from the web to use in calculations.
So I'm not sure that I would describe OpalCalc as "LastCalc on steroids" by any stretch.
It's software, the website is just one place you can obtain the software but there are others.
I'm fairly sure that merely using LastCalc (ie. being a user of the web service) doesn't impose any responsibilities.
Ok, just for you I risked borking the site during a slashdotting and I implemented a quick fix. You're welcome :-)
Yeah, you're confusing it with a recursive function definition, I've been meaning to fix that. I guess I'll fire up Eclipse (it's Java, not Lisp)
This is compelling but the use of Affero for the license makes onerous demands of the user. The implicit threat of a code audit is there.
Can you elaborate? Which clauses specifically make onerous demands?
I'm guessing you haven't heard much about British libel laws.
Why are they coming up with their own operating system and app ecosystem, is this really the core competency of a car company? Why aren't they using Android, which already has text to voice, voice to text, GPS navigation, and almost everything else you might need in a car?
Slashdot no-longer looks like a car crash by reverting to essentially the level of site complexity they had in 2006 (or, at least, disguising the "improvements" more effectively). Will be interesting to see if this stems or reverses the exodus of readers /. has experienced over the past half-decade.
The ad model is working extremely well for Google, Facebook, and others.
Rather it is an auto-immune disorder, gluten causes the sufferer's own immune system to attack their small intestine. Aside from the immediate pain (a stomach ache for a day or two after eating even a tiny amount of gluten), it can result in deficiencies in various necessary substances, and can lead to an increased danger of cancer. My wife had stomach aches most of her life, she had grown accustomed to them, thinking they were normal. A few years back (she was 28), on someone's suggestion she got tested for Celiac (first a blood test, then a biopsy of her small intestine). She was positive. She has been avoiding gluten ever since, she can't even have a single crumb of bread without getting sick now. Most people that have Celiac are never diagnosed, and suffer a life of pain and misery as a result, in addition to a shortened lifespan. If you get a lot of stomach aches for reasons you can't determine, you owe it to yourself to get a blood test for Celiac.
And yet I'd bet you'd be the first to expect to get treated in an emergency room if you didn't have health insurance and something bad happened. The reality is that nobody wants to live in a country where people are allowed to die on the streets because they are poor. If you accept that as a premise, then at some point it is necessary for us all to accept some kind of mandate to participate in the health care system. The alternative is the current situation, the worst of all worlds, where emergency rooms end up being the safety net for those without insurance, and we all pay through the nose for it.
One reason this kind of problem occurs is that many collaborative filtering algorithms are measured based on "root mean squared error", basically the square root of the mean of the differences between what was predicted and what the user actually did.
The problem with this metric? It doesn't account for a variety of important things, one of which is that most users value diversity. Another is that in most recommendation systems, what is important is the relative relevance of recommendations to each-other, whereas RMSE is an absolute measure of effectiveness. And a really tricky one is that the recommendation algorithm itself can impact user behavior. For example, the user may raise their standards if the algorithm does a better job.
The unfortunate answer is that the only rock-solid way to measure the effectiveness of recommendation algorithms is to test them with real users, perhaps splitting the user population between different algoritms, and seeing which does best.
I'm pretty familiar with this issue as my day job is building a behavioral ad targeting engine. We learned a long time ago that while RMSE has its uses, there is often limited correlation between an algorithm's ability to predict user behavior retrospectively (which ads they will click on and what products they will buy), and how much additional revenue the algorithm will generate in practice.
Our solution is to use RMSE as a first-blush indication of how good an algorithm is. Secondly, we take the top, say, 10% of ads with the best predictions, and see what the actual click or conversion rate is within this 10%. This requires a higher volume of data, but yields results that are closer to what we find in reality. Lastly, the algorithm then has to prove itself in the wild on a small subset of traffic. Only then can we really know if any algorithm is an improvement on any other.
Freenet has been slow and hard to use in the past, but its improved quite a bit. It is the obvious platform for something like Wikileaks. Of course, there is nothing to prevent people from mirroring content on the web (since installing Freenet, like any piece of software, is a hassle). But at least there will be an unimpeachable backup of all data on Freenet.
You are an idiot. If anyone is worse than Microsoft it is these patent trolls.
Its unsolvable to do it in advance, but quite possible to do it while observing the running code (a bit like how a filesystem optimizes the locations of data on disk).