Forgot your password?
typodupeerror

Comment: Re:Just how powerful *IS* faith? (Score 1) 217

by EPAstor (#43319075) Attached to: Interviews: James Randi Answers Your Questions

I'm not saying that what you're saying isn't possible, and could very well be what happened, but that doesn't change the WTF factor when something is supposedly not going to go away without surgery.

Well, yes. The issue here is "supposedly". I've had a bit of experience dealing with the medical system, and (at least in the US) there's an epidemic of overconfidence to the point of idiocy among doctors. This manifests in several ways, such as:

1) If the tests that ONE DOCTOR thinks are worthwhile come back negative: "There's nothing wrong with you, it's all in your head." Even if there are significant possibilities still not excluded. (Note: in any remotely complicated presentation, it's extremely rare that two doctors will think of the same set of tests to run... and many doctors don't actually know the proper best-practices for testing for various conditions, even within their own specialty.) On the flip side of the issue - if a test comes back positive: "That's the issue!", even in cases where false positives are incredibly common. This is overridden only in very specific circumstances, where the doctor is trained to question a specific test as routinely giving false positives - their statistical training is often shallow enough that they cannot extend this past the explicit guidelines.

2) Once a doctor decides that they can treat something, they will often exaggerate the need for treatment, saying things like "This will not go away without surgery" even in cases where as much as 10% of such phenomena do. Or more. Think about how NEW physical therapy is as part of accepted medical practice... and how many illnesses are best treated by physical therapy, where doctors had always resorted to surgery before. Chronic lower back pain has now been shown by scientific study to have essentially identical outcomes by surgery and by physical therapy - yet, if you asked a surgeon whether the surgery was beneficial, they always said yes. The best then went on to point out that the added risk and complication might balance out the benefits - but they always held that there was a significant improvement in the best outcomes.

3) Confirmation bias. Once a doctor or technician highlights an abnormality, it's remarkable how many people will find the same thing when looking at the data - even though, provided with the data and NO information on the diagnosis, they would not have found the issue. You know how scientific experiments are usually held to the "double-blind" standard? This is specifically invented to limit confirmation bias - because it has been shown, over and over again, to be incredibly powerful. In a noisy image, we tend to see what we expect to see.

In other words - spontaneous remission is possible. So is a mistaken diagnosis, even with confirmation, unless you have reason to believe that the second technician read the data blind to the previous technician's work. Personally, I suspect that the surgeon strongly exaggerated their case - as they often do. "This NEVER goes away on its own" may mean that the likelihood is less than 10%. Or less than 1%. Or even 1 in 10,000! Unless the disease is extremely rare (in which case any such statistics are questionable due to a small sample size), 1 in 10,000 still means that it happens to plenty of people in the United States alone.

Comment: Science Fiction, Anyone? (Score 2) 315

by EPAstor (#41632083) Attached to: DRM Could Come To 3D Printers
Huh. This is DEFINITELY one of the cases where anyone who reads Sci-Fi knows there's prior art, in the sense of published material describing a system operating in essentially this way. Patent was filed on January 31, 2008... Anyone want to help find stories that mention volumetric printer DRM pre-2008? Cory Doctorow's used the point in several stories - but Makers, at least, wasn't published until 2009. Anything pre-dating? Also, I think I've read an old classic short story that described people surviving a war by use of a synthesis device where they'd disabled the mechanism that prevented the creation of various goods... anyone know what I'm talking about?

Comment: Re:Hoping to Clarify ... (Score 5, Interesting) 730

by EPAstor (#39168391) Attached to: YouTube Identifies Birdsong As Copyrighted Music

Mr. Bakke, please explain how submitting the note "All content owners have reviewed your video and confirmed their claims to some or all of its content" is possibly acceptable when no one's reviewed it? This makes your company look awfully bad.

Alternatively, if someone did review it and sign off on that reply, then I hope this will reflect properly on that individual and their career at Rumblefish, as I'm pretty sure that this at least makes your company rather vulnerable from a publicity point of view, if not a legal one. If you're outsourcing this... then really, I hope your company can learn its lesson QUICKLY.

Comment: Re:Such systems have been proposed before (Score 2) 1065

by EPAstor (#38979107) Attached to: The Zuckerberg Tax
The problem is that sales taxes are inherently regressive, not flat. It sounds ridiculous, I know. But - if you live paycheck-to-paycheck, then by definition, you're spending all of your money buying things. If you have income high enough that you're putting away savings, then you spend a smaller fraction of your money - and thus a smaller fraction of your money is subject to the sales tax. Therefore, a nominally flat sales tax actually taxes the poor more harshly, percentage-wise, because the poor find it necessary to spend a larger fraction of their money. Same logic applies to almost all consumption taxes, save specific luxury taxes (which I disapprove of on entirely different grounds).

Comment: Re:Of course it was possible (Score 2) 212

by EPAstor (#38501050) Attached to: What If Babbage Had Succeeded?

Imagine how complex, expensive, and precise the machinery needed to perform WWII-era ciphers would be if it were purely mechanical. It would also have to be fairly single-purpose.

You mean, like the Enigma machine? Remarkably simple code... breakable with sufficient processing, or with improper use of the protocols, it's true. However, for all intents and purposes it required a highly-specialized bank of mechanical computers to break the code, and it generally took quite a while if the encryption was being used properly.

Comment: Re:Such an awesome crowdsourcing success! (Score 4, Informative) 417

by EPAstor (#37508342) Attached to: AIDS Vaccine Breakthrough
Wrong breakthrough, I'm sorry to say. That one was an analysis of a protein that all retroviruses (including HIV) have - this one is an actual (albeit in vitro) treatment method. This paper is in a completely different direction, and arguably one step further along its path... and no, FoldIt was not involved in this particular breakthrough. Both are cool, but not the same work.

Comment: No Infringement (Score 1) 182

by EPAstor (#36501648) Attached to: Bittorrent and uTorrent Sued For Patent Violations
IANAL, but I don't get it. On a brief reading of the claims, and contrasting them with what I know of how BitTorrent works, I can't see how BitTorrent violates any of the Claims. Specifically, all of their claims include Claim 1, which is as follows:

Claim 1: A media distribution system, comprising:
a media file database configured to store media files, wherein one or more of the media files have been compressed prior to storage in the media file database;
a computing device configured to receive user requests for delivery of the one or more of the media files stores in the media file database, the computing device further configured to: (identify average network throughput between computing device and the requesting users; and route the user requests for delivery of the requested one or more media files to a distribution server capable of servicing the user requests based upon at least the average network throughput;) and
a distribution server coupled to the media file database, the distribution server configured to simultaneously deliver a single copy of the requested one or more of the media files identified in the routed user requests to the requesting users in less-than-real-time, wherein the distribution server automatically adjusts delivery of the requested one or more media files to the requesting users based on current average network throughput between the distribution server and the requesting users.

That isn't quite BitTorrent. Specifically, I don't think BitTorrent shapes its routing from the server-side based on "average network throughput between computing device and the requesting users". Nor is the system attached to a database configured to store media files - at least, I hope filesystems in general don't count.

Computers will not be perfected until they can compute how much more than the estimate the job will cost.

Working...