Cap'n Refsmmat

Want to share the nerdly goodness of ScienceForums.Net with your Facebook friends? Go right ahead! As of today, it's now possible to connect your SFN profile with your Facebook and Twitter profiles -- and if you haven't yet registered on SFN, you can use Facebook Connect or your Twitter account to sign up in half the time. If you'd like to try it out, head over to your Facebook Connect settings or Twitter Connect settings and follow the instructions. After a few steps, you'll be able to connect your SFN profile to your Facebook or Twitter profile: Note that all these features are optional, and you can turn them off if you'd like. Either way, you'll be able to log in to SFN by clicking the Facebook or Twitter icon on the upper right, so you have to remember one fewer password. Once you're connected, you can share posts and topics on Facebook and Twitter, display your Facebook and Twitter status updates on your SFN profile, and much more. And while you're at it, be a fan of SFN! @ScienceForums Note for the paranoid As per our privacy policy (section 6), we will not share your information with anyone unless you specifically allow it. At no stage in the process does SFN receive your Facebook password, and we will not post content on your Wall unless you enable the settings or do it yourself.

Okay, I updated it a bit, with a link to the new chat. There might be more revisions to do, but I just got it functional again.

It's fine to post this kind of thing in Computer Science if you want to discuss the computer science aspects of it, so I have no issue there. Whether or not computer science is a "real" science doesn't really matter to me, although you could certainly open a discussion on that...

Yeah, I was greatly amused when I spotted it. The majority of my Junk folder is "Your manhood will get support here!", "Perfect tab for wang strength," and "Rolex.com For You -63%". I'm currently struggling to set up automatic deletion -- this goes back to April 15th and it's 8000+ mails.

So we operate a staff at scienceforums dot net email address for official contact. As of right now, it has 8654 junk mail messages sitting around, and I was going to start deleting when I noticed this beautiful email: Best email ever.

Bail has been granted and the extradition hearing will be January 11. http://www.guardian....sange-wikileaks I expect we'll see announcements from Assange as soon as his lawyers have come up with the money and he's released.

That's not entirely true. There were 15,000 documents withheld from the Afghan war logs because they may harm people. The trouble was that some of the other documents should have been withheld as well.

I'm afraid that's slightly above my pay grade.

I might also spend some time trying to make stuff. Like, say, this: http://makeprojects.com/Project/See-Thru-Potato-Cannon/5/1 But clearly that's a few days of project at most. What fun projects could I sustain all winter break?

How so? Wikileaks doesn't provide you the means to steal secret documents and commit the illegal act. Assange, as far as I know, hasn't sent anyone hidden cameras or special encrypted thumb drives. He provides means to publish documents, which in itself is not illegal (or at least tenuously illegal). I suppose you could say Wikileaks provides anonymity and aids leakers in avoiding arrest. Like a service that carefully disposes of bodies so you can't be caught after committing murder. Hm.

Has it been established that Wikileaks breaks the law?

I thought I said that the NFL theorems don't apply to natural selection as it occurs in the wild. Since you agree that the paper I cited agrees with Dembski and Marks, you'll also have to agree that Dembski and Marks' work can't apply to a fitness function that can change independently of time. There's a difference between "primary cause" and "supplying information for the design". I asked you to cite specific conclusions. But never mind, because I'll be rather busy reading for a while and may not be able to respond. I'll get back to you once I've read all those papers.

Please see D. Wolpert and W. G. Macready, "No free lunch theorems for optimization," IEEE Trans. Evolutionary Computation, Vol.1, No.1, pp. 67-82, 1997, as cited by Demsbki. Specifically, page 68, which states that the work is limited to a "cost function" or "objective function" to be optimized (also known as a "fitness function" in our application) that is either static or time-dependent. Functions that vary depending on the current location in the search space are not included. There are many open hypotheses for this question. One potential answer is that time and an abrupt change in the fitness landscape (such as a change in the environment or in competitors) forced adaptation to proceed in a certain direction. Annu. Rev. Earth Planet. Sci. 2006.34:355-384 provides an interesting overview of the possibilities in fitness landscape change in one particular event, but it is by no means an exhaustive report on this subject. I said "not bounded." cite [sayht] to quote (a passage, book, author, etc.), esp. as an authority: He cited the Constitution in his defense. By specific, I meant "list the actual peer-reviewed papers that support this particular conclusion, rather than a list of hodgepodge book chapters, conference proceedings and actual papers that cover a variety of topics, many of which don't cover the exact statement I asked you to cite." Fair enough. We'll resume this discussion once I've read them all.

Edtharan didn't specify the size of his box. A larger size makes the scenario correspondingly likely, but not physically impossible. In any case, my argument has nothing to do with processes occurring over an extended period of time. It has to do with Edtharan's example. If his example is faulty, that's not my fault.

Indeed. And Edtharan said "There is no law of physics against this occurance.... just that it is highly unlikely to occur." Probability. I wasn't arguing that the Second Law is wrong. I was arguing what your link argues, and what Edtharan said: It is statistically highly unlikely that what Edtharan says will occur, but not impossible. That is the point.

I said "Specifically" because I knew you would cite this. Many of the papers there are actually book chapters or less-formally-reviewed conference proceedings, and quite a few deal specifically with the success of search algorithms, not the generation of information. Could you be more specific? This paper seems to reject the definition of "information" used by most of information theory, so it's hard to grasp its relevance to our current conversation or how it fits into any other presented research: To support your claim of published papers in information theory (not search theory, or a new system that rejects information theory) you'll have to produce specific papers showing how evolutionary algorithms based on physical systems cannot create adequate information. A paper titled "The failure of evolutionary algorithms to generate functional information" would be rather nice, if you have it.

No, it isn't. See the fluctuation theorem, and, for example: G.M. Wang, E.M. Sevick, E. Mittag, D.J. Searles & Denis J. Evans (2002), Experimental demonstration of violations of the Second Law of Thermodynamics for small systems and short time scales, Physical Review Letters 89: 050601/1–050601/4. doi:10.1103/PhysRevLett.89.050601 This is particularly important for very small boxes.

An evolutionary model can be used to explore the current diversity of life and explain why various species exist in various places, testably. Does design offer a testable similar hypothesis? How can a designer explain If you would read the paper, you would see that search algorithms are capable of generating information. See section 4, which points out that "a good search is one that generates the active information necessary for success" and explains that for such a search to succeed, it must either take an exceedingly long time, or be given information about the search space. That is, blind searches can generate information, just very slowly. Furthermore, I will note yet again that the NFL restrictions do not apply here. As for my conclusions about natural selection's endogenous information: it's quite simple. Suppose we have a multidimensional search space and our blind walk algorithm must traverse all of it to find an area of maximal fitness. A totally random walk would take an exceedingly long time. However, natural selection provides some information about the fitness space. For example, if the random walk tries to walk into a large region of low fitness, it will be forced to stop -- the organisms mutating in this direction will die, and will not explore that portion of the space. On the other hand, organisms randomly walking into a higher-fitness portion of the search space will thrive. By having more offspring, they will increase the number of organisms exploring this part of the search space, and will naturally explore it faster, by virtue of being more fit and having more offspring. That's not what I said. The rate of production is limited by the information input, but the total amount is not bounded by that. Unless you wish to provide a peer-reviewed citation stating otherwise? So: random processes can create information in small amounts, but as you said earlier, they can't create any information? You will kindly cite this claim. I provided the design documents, which show that the evolutionary algorithm merely measures whether the antenna will match the propagation characteristics desired for the probe. The fact that it designed an antenna more functional than those created by a human designer undermines the claim that human designers made it. The part saying "pre-existing information of equal or higher order was required". Cite it.

Fair 'nuff. Palin says he should be "hunted down", and there was that Canadian former government advisor who openly called for assassination. The politicians have merely screamed for prosecution, though Huckabee called for the execution of whoever leaked the documents. You neglected to mention the insurance file, which purportedly has all the unreleased cables. Also, the five newspaper partners have copies of all the unredacted cables, so an attack would have to take out the news sources as well -- which would be far more dangerous than attacking Wikileaks. (Newspapers enjoy better PR and more lawyers.)

That doesn't answer the question. Is there a testable model for how intelligent design accounts for the current diversity of life? Could you cite these? Specifically. This published paper disagrees with you: http://evoinfo.org/p...h-for-a-search/ It shows that that evolutionary algorithms (and other simple searches) can generate information at a rate proportional to the amount of basic information available about the search space. Natural selection of course has great knowledge of the search space, since it is physically limited from exploring harmful parts of the search space. The search algorithms in this paper are in fact more limited than natural selection is in nature, because they are constrained to a fitness function that does not change in response to new adaptations. (Else the NFL theorems wouldn't apply.) Furthermore, if I supply 10 bits of information to a search algorithm as described in the Dembski paper, it can generate more information at a rate determined by the information I supplied. A totally blind search is not prevented from generating information either, except in the sense that time constraints make it impractical. If I recall correctly, I cited an antenna example and asked you to provide evidence that supplied information caused the final design, and you never came up with any. Can you cite a journal article supporting this claim?

We don't have to limit ourselves to conspiratorial allegations of CIA involvement in the rape charges. The fact that prominent politicians demanded Assange's execution, abduction or assassination is alarming by itself. Then there are the claims that Wikileaks is a "terrorist organisation" or a "transnational threat" that needs to be hunted down and stopped. In a country like ours, we should not be advocating that the government freely break its own laws to hunt down someone who possibly didn't even break the law at all. If only we had some sort of mechanism to have a group of people decide if the charges against a person are valid, and an appropriate punishment then be handed out to the perpetrator...

This'd require that your ISP log your traffic in the same detail as the target site, or that someone be watching your connection in real time. Furthermore... I could just use an encrypted VPN over Tor, so attacks would be traced back to the VPN server, and the VPN couldn't be traced back to me. You couldn't correlate the VPN connection to Internet activity since the connection would be continuous, so the ISP would record one connection and you'd do numerous things through that connection. You could also add random scrambling activity (visiting random websites, sending random data, etc.) to mess up correlations. So... the government can do things to catch you, but you can do things in response. The nerds always have one more trick. Indeed. I'm sure these events will spur further development of Tor/Freenet-like systems. We'll just have to find out... Ah, something interesting from the latest cables: http://www.guardian.co.uk/business/2010/dec/09/wikileaks-cables-pfizer-nigeria From the cable itself: Pfizer says:

Tor has some specific attacks against it, yes, but not necessarily anonymity-busting ones. You can sniff traffic coming out of an exit node, but that won't give you the identity of the user, just whatever sites they're visiting. You can do various sniffing attacks on the rest of the network and narrow down the possibilities, but AFAIK there's no way to listen and break the anonymity totally. As for Freenet, careful use of darknets can mitigate the problem, I think. I don't know much about I2P. My point is that the nature of the Internet makes it very difficult to police. You can only catch people using Tor if you're watching at exactly the right time -- if you begin investigating after the fact, you won't be able to mount the surveillance required to narrow down the users. Systems like BitTorrent make it very difficult to stop the distribution of files once it has started: you can try to shut down the central server, but once a few people have the file, they can share it with everyone else. The widespread availability of high-grade encryption makes it easy to hide secrets from the government. Your only option would be to completely remake the Internet with security and identity features built-in.

Low blood pressure upon standing is orthostatic hypotension, and I encounter that very frequently myself -- but there's no shaking or hyperventilation. I merely stand up, feel light-headed, and experience dimmed vision. Otherwise, I agree with D H. At least get your blood pressure checked to see if it's chronically low.