tag:blogger.com,1999:blog-13030925.post1176797276307470644..comments2024-01-25T13:46:11.967-06:00Comments on The Bronze Blog: A Little StoryRyan Michaelhttp://www.blogger.com/profile/14750814560493466382noreply@blogger.comBlogger5125tag:blogger.com,1999:blog-13030925.post-46511713804967875562010-06-11T05:27:38.264-05:002010-06-11T05:27:38.264-05:00WOAH! There's so many layers of awesome in th...WOAH! There's so many layers of awesome in that article I don't know what I should be more amazed by! The genetic algorythm-on-a-chip? The genetic algorythm for making a better GENETIC ALGORYTHM? The potential for developing AI (which, as a side effect, we STILL wouldn't be any closer to understanding, but would at least have better tools to reverse engineer). Or, is it the idea of a chip that can actually rewire itself into other forms? I'm sure there's some limits, like how many semiconductor gates it can form at once, but think of this. Emulating computer systems would be a thing of the past. Program in a specific computer profile and you could get a perfect circuit replica of a Super Nintendo. That's awesome.<br /><br />More to the point, it's an example of how scientists and engineers actually DOING work are infinitely more interesting than the quacks who can do nothing but sit around making stuff up.Dark Jaguarnoreply@blogger.comtag:blogger.com,1999:blog-13030925.post-88071101897400932010-06-10T06:42:27.557-05:002010-06-10T06:42:27.557-05:00Oh yeah, genetic algorithms are hoaxes - that'...Oh yeah, genetic algorithms are hoaxes - that's why it's perfectly normal for them to produce results that nobody can understand, but work anyway. <i>Sure</i>.<br /><br />"Adrian Thompson at the University of Sussex in the UK ... made a genetic algorithm which tested various configurations of the [FPGA] chip so that it would generate a 1 volt signal if it detected a 1-kilohertz audio tone and a 5 volt signal if it detected a 10-kilohertz audio tone. After a certain amount of evolution, the program worked brilliantly, but what is downright scary is this: the FPGA only used 32 of its 100 available logic gates to achieve its task, and when scientists attempted to back-engineer the algorithm of the circuit, they found that some of the working gates were not even connected to the rest through normal wiring. Yet these gates were still crucial to the functionality of the circuit. This means, according to Thompson, that either electromagnetic coupling or the radio waves between components made them affect each other in ways which the scientists could not discern (Taubes 1997)."<br /><br />http://biology.kenyon.edu/slonc/bio3/AI/GEN_ALGO/gen_algo.htmlDuncnoreply@blogger.comtag:blogger.com,1999:blog-13030925.post-90395930018581462112010-06-10T05:00:53.216-05:002010-06-10T05:00:53.216-05:00They handwave that as some ridiculously unrealisti...<i>They handwave that as some ridiculously unrealistic expectation... Annoying...</i><br /><br />Although they fully expect you to research every single little crappy anecdote they through at you.Valhar2000https://www.blogger.com/profile/05467019327257867276noreply@blogger.comtag:blogger.com,1999:blog-13030925.post-43939844445048039252010-06-09T19:20:19.002-05:002010-06-09T19:20:19.002-05:00Creationists who don't know much about program...Creationists who don't know much about programming offer up some very annoying responses to genetic algorithms. Their main arguments are just different forms of "computers are man-made so they don't count". At no point in any form of it do they actually explain why the designed nature of the computer would somehow imbue the blind selective nature of the program with immunity from their criticism that blind selection is impossible. It's just some magical contaminant that poisons everything along the way or something.<br /><br />Others make a more annoying argument, that "genetic algorithms" are hoaxes. Most of them have never even HEARD of them until I bring it up. I ask them to look at the code, but then they come up with that defense of "do you honestly expect me to research EVERY LITTLE THING out there?", and my response is simmply "if you're going to make a claim to knowledge about something, yes, I expect that". They handwave that as some ridiculously unrealistic expectation... Annoying...Dark Jaguarnoreply@blogger.comtag:blogger.com,1999:blog-13030925.post-52428896918741970282010-06-09T03:46:48.914-05:002010-06-09T03:46:48.914-05:00Man, I find genetic algorithms absolutely fascinat...Man, I find genetic algorithms absolutely fascinating. Some of the results they come up with are so original and work so well that they defy belief.Valhar2000https://www.blogger.com/profile/05467019327257867276noreply@blogger.com