I do have one question for them, though. If they can't afford to do Internet-based television, what makes them think they can afford to run a space program? I'm just curious, that's all.
Perhaps because they think that instead of spending loads of money on improving their bandwidth infrastructure for an unusually large two-time spike, it'd be better to spend that money on science and exploration?
Try scrolling down a little on the MSNBC page, to the "This space for rent" subtitle. I tried including an anchor in the link that goes down there, but it doesn't seem to be working right. I think the # character is getting converted to something else (can't verify that myself right now, on Treo, but will later).
Any information broadcast into cortex would be very crude and likely also unpleasent.
Curiously enough, this is an accurate characterization of many sources of popular entertainment. Brindley's device may not have worked out as a long-term visual prosthetic due to its crude input properties, but if a computer-controllable device with similar crude capabilities were available, safe, and noninvasive, I guarantee that they'd sell like crazy.
Unless I'm grossly mistaken, haven't there already been studies which used transcranial magnetic stimulation (TMS) to induce phosphenes in sensory cortex areas like V1 and MT? Granted, targetting is still quite crude, but I don't see any particular reason why the resolution can't be improved in future techniques. I don't we're going to be stimulating individual cortical microcolumns non-invasively any time soon, but it certainly seems possible to put together something which would be useful enough for entertainment purposes. Getting FDA (or whatever the associated gov body is) approval though is an entirely different question.
In any case, ultrasound seems like it might be an interesting research tool for neural stimulation. Unfortunately, I'm not familiar enough with they physics though to know how feasible it is.
A Treo 600 or 650 with iPedia works nicely for this and has a well-designed interface. It's shareware, but IMHO well worth it -- I find myself using it quite regularly.
Coincidentally, solar satellites constructed from in-space resources have often been touted as a possible solution to our oil dependency. As another responder stated, devoting all resources to "problems here on earth" while ignoring everything else generally isn't the best way to go.
The Bush Administration has majorly cut back NASA's budget, leaving them with little choice.
Hating Bush because he's Bush is cool and all, but he actually increased NASA's budget. They got a 5.6% increase for 2005 and are expecting (maybe already got?) a 4.7% increase for 2006. NASA is one of the few non-defense arms of the government which has gotten a funding boost, and there's actually worry that it might make it a target for other interests ("We should spend money on earth," blah blah).
This is from last month, but Space Daily's Bruce Moomaw has an extensive overview of NASA's future plans for Mars exploration, based on the results of the first meeting of the Mars Strategic Roadmap Committee. It's a highly recommended read.
Some highlights: * The 2007 Phoenix will "land on the near-surface layer of ice-saturated ground discovered by the Mars Odyssey orbiter in Mars' north polar regions to study the ice itself and its potential for preserving biochemicals." * Mars Telecommunications Orbiter in 2009, which could boost the data rate coming back from Mars 10x to 100x. * The Mars Science Laboratory will likely be pushed back to 2011 (instead of 2009), but is likely to have two or more versions constructed and sent to different areas. The base cost for a single rover is estimated at $1 billion, but another rover is expected to add $400 million. The MSL (or MSLs) will be looking for traces of organic chemicals and be further investigating the geological/climate history of Mars. The MSL is expected to weigh 600 kg including 65 kg of scientific instruments, compared to the MERs which weigh 185 kg including 5 kg of scientific instruments. * There still seems to be considerable debate over when and how to launch a Mars Sample Return mission. One proposal I like is to send one (or more) to land near a MSL, have the MSL load a pre-drilled soil sample into the MSR, and then have the loaded MSR's return vehicle launch back.
The question of whether it is possible to automate the scientific process is of both great theoretical interest and increasing practical importance because, in many scientific areas, data are being generated much faster than they can be effectively analysed. We describe a physically implemented robotic system that applies techniques from artificial intelligence to carry out cycles of scientific experimentation. The system automatically originates hypotheses to explain observations, devises experiments to test these hypotheses, physically runs the experiments using a laboratory robot, interprets the results to falsify hypotheses inconsistent with the data, and then repeats the cycle. Here we apply the system to the determination of gene function using deletion mutants of yeast (Saccharomyces cerevisiae) and auxotrophic growth experiments. We built and tested a detailed logical model (involving genes, proteins and metabolites) of the aromatic amino acid synthesis pathway. In biological experiments that automatically reconstruct parts of this model, we show that an intelligent experiment selection strategy is competitive with human performance and significantly outperforms, with a cost decrease of 3-fold and 100-fold (respectively), both cheapest and random-experiment selection.
I'm fairly certain that Ray Kurzweil and Hans Moravec at least know of each other, having served together on various panels. I wouldn't be surprised if the charts were used with permission. Considering that the overlap of their readership is pretty large, blatant plagiarism would be pretty easy to catch.
One of the bigest problem with neural networks is that 99.99% of all implementations are linear.
Do you have something to cite for that statement? I'm under the impression that just about everybody doing things with neural networks incorporates at least some sort of nonlinearity. Sigmoid functions and radial basis functions seem most popular nowadays, with the occasional spiking neural network thrown in to spice things up.
My personal opinion is that there was way too much of an emphasis on supervised learning in the 80s and 90s -- basically just finding ways to match up inputs and outputs. IMHO, unsupervised learning is where the fun stuff is.
This reminds me a little bit of a rather neat system I came across the other day, Video Google (despite the name, I don't think it has anything to do with the Google company). It doesn't use metadata or cheats like that, but rather uses image analysis to identify recurring objects and scenes.
They have a demo on their web site where you can select a portion of a video frame, and it'll show you all the places in the movie where the algorithm thinks that snippet shows up. Some other cool examples are displaying the appearances of a clock from 'Groundhog Day," and a recurring poster from 'Run Lola Run.' A research paper with more details is available here.
The abstract:
We describe an approach to object and scene retrieval which searches for and localizes all the occurrences of a user outlined object in a video. The object is represented by a set of viewpoint invariant region descriptors so that recognition can proceed successfully despite changes in view-point, illumination and partial occlusion. The temporal continuity of the video within a shot is used to track the regions in order to reject unstable regions and reduce the effects of noise in the descriptors.
The analogy with text retrieval is in the implementation where matches on descriptors are pre-computed (using vector quantization), and inverted file systems and document rankings are used. The result is that retrieval is immediate, returning a ranked list of key frames/shots in the manner of Google.
The method is illustrated for matching on two full length feature films.
Hating Bush is all well and good, but could you elaborate more on your space robotics -> militarization connection? You seem to be stating it rather axiomatically, and I'm afraid I don't follow your logic.
Disclaimer: I'm an avid roboticist and have done robotics research in the past. I'm fairly neutral on militarizing space, but am a very big advocate of developing robotic technologies, particularly because they're damned useful for pretty much any space-related activity.
Not to nit-pick, but Bush has actually increased NASA spending. It's one of the few non-defense portions of the government which have actually gotten increased funding under his watch.
I already use programs like TiBR to read textfile novels on my Treo 600. It's a little scrunched, but I don't really notice it; things are probably considerably better on the Treo 650.
Project Gutenberg has plenty of textfile novels ripe for reading. Cory Doctorow's stuff is also pretty good. I read his Down and Out in the Magic Kingdom entirely on my cell.
Another handy resource is the University of Virginia Library's Etext Center, which has 1800 freely-available eBooks.
I think a better option than repair would be putting together the spares on the ground, including a PERFECT mirror as opposed to the flawed one that got launched, and launch Hubble II.
I agree that this would be good, although I doubt very much that it can be done by cobbling together leftovers. Last I saw, the estimated cost of such a thing was around $4B, vs. under $2B to fix the one we've got.
Actually, the cost of a new and much improved Hubble (including launch) would be more around $1B:
Nobody ever seems to mention the Hubble Origins Probe in these discussions, which is IMHO the best possible solution:
An international team led by Johns Hopkins University astronomers have proposed an alternative to sending a robotic or manned repair mission to the ailing Hubble Space Telescope. Their proposal is to build a new Hubble Origins Probe, reusing the Hubble design but using lighter and more cost-effective technologies. The probe would include instruments currently waiting to be installed on Hubble, as well as a Japanese-built imager which 'will allow scientists to map the heavens more than 20 times faster than even a refurbished Hubble Space Telescope could.' It would take an estimated 65 months and $1 billion to build and launch, approximately the same cost as a robotic service mission.
The original Hubble is great for historical and sentimental reasons, but the cost/benefit ratio is really so much better with a replacement.
Another point: Would you want to add an additional condition that all awardees would have to release the source code to their work, to allow future contestants to build on it? I believe the RoboCup robotic soccer competition does something like this.
The only problem with it is that it doesn't include Mahoney's threshold of 1.3 bits per character for artificial intelligence.
Indeed, but I don't see this as much of a problem. It favors steady incremental progress, which should eventually surpass the threshold.
Hehe. In any case, I think my Karma's been maxed out for a years now. If it helps my meta-karma, though, I've also contributed to some of the articles which I linked to.
Slashdot Mirror
I had better not find myself jaywalking next time I go to Otakon, or else I might get shot on sight.
Hah! Recent studies show that Rikku outfits and Con badges are clear indicators of terrorist activity.
I do have one question for them, though. If they can't afford to do Internet-based television, what makes them think they can afford to run a space program? I'm just curious, that's all.
Perhaps because they think that instead of spending loads of money on improving their bandwidth infrastructure for an unusually large two-time spike, it'd be better to spend that money on science and exploration?
Try scrolling down a little on the MSNBC page, to the "This space for rent" subtitle. I tried including an anchor in the link that goes down there, but it doesn't seem to be working right. I think the # character is getting converted to something else (can't verify that myself right now, on Treo, but will later).
Any information broadcast into cortex would be very crude and likely also unpleasent.
Curiously enough, this is an accurate characterization of many sources of popular entertainment. Brindley's device may not have worked out as a long-term visual prosthetic due to its crude input properties, but if a computer-controllable device with similar crude capabilities were available, safe, and noninvasive, I guarantee that they'd sell like crazy.
I'm sort of a neuroscientist...
Unless I'm grossly mistaken, haven't there already been studies which used transcranial magnetic stimulation (TMS) to induce phosphenes in sensory cortex areas like V1 and MT? Granted, targetting is still quite crude, but I don't see any particular reason why the resolution can't be improved in future techniques. I don't we're going to be stimulating individual cortical microcolumns non-invasively any time soon, but it certainly seems possible to put together something which would be useful enough for entertainment purposes. Getting FDA (or whatever the associated gov body is) approval though is an entirely different question.
In any case, ultrasound seems like it might be an interesting research tool for neural stimulation. Unfortunately, I'm not familiar enough with they physics though to know how feasible it is.
This publication from 1995, Application of focused ultrasound for the stimulation of neural structures seems interesting, but I don't have access to the fulltext.
I tried asking Google if entropy could be reversed, but it seems that there's insufficient data for a meaningful answer. ;)
A Treo 600 or 650 with iPedia works nicely for this and has a well-designed interface. It's shareware, but IMHO well worth it -- I find myself using it quite regularly.
Woah, didn't those things show up in Metal Gear Solid 2, as hover sentries or something?
Coincidentally, solar satellites constructed from in-space resources have often been touted as a possible solution to our oil dependency. As another responder stated, devoting all resources to "problems here on earth" while ignoring everything else generally isn't the best way to go.
The Bush Administration has majorly cut back NASA's budget, leaving them with little choice.
Hating Bush because he's Bush is cool and all, but he actually increased NASA's budget. They got a 5.6% increase for 2005 and are expecting (maybe already got?) a 4.7% increase for 2006. NASA is one of the few non-defense arms of the government which has gotten a funding boost, and there's actually worry that it might make it a target for other interests ("We should spend money on earth," blah blah).
This is from last month, but Space Daily's Bruce Moomaw has an extensive overview of NASA's future plans for Mars exploration, based on the results of the first meeting of the Mars Strategic Roadmap Committee. It's a highly recommended read.
Some highlights:
* The 2007 Phoenix will "land on the near-surface layer of ice-saturated ground discovered by the Mars Odyssey orbiter in Mars' north polar regions to study the ice itself and its potential for preserving biochemicals."
* Mars Telecommunications Orbiter in 2009, which could boost the data rate coming back from Mars 10x to 100x.
* The Mars Science Laboratory will likely be pushed back to 2011 (instead of 2009), but is likely to have two or more versions constructed and sent to different areas. The base cost for a single rover is estimated at $1 billion, but another rover is expected to add $400 million. The MSL (or MSLs) will be looking for traces of organic chemicals and be further investigating the geological/climate history of Mars. The MSL is expected to weigh 600 kg including 65 kg of scientific instruments, compared to the MERs which weigh 185 kg including 5 kg of scientific instruments.
* There still seems to be considerable debate over when and how to launch a Mars Sample Return mission. One proposal I like is to send one (or more) to land near a MSL, have the MSL load a pre-drilled soil sample into the MSR, and then have the loaded MSR's return vehicle launch back.
This reminds me of a Nature paper from last year:
Functional genomic hypothesis generation and experimentation by a robot scientist
The question of whether it is possible to automate the scientific process is of both great theoretical interest and increasing practical importance because, in many scientific areas, data are being generated much faster than they can be effectively analysed. We describe a physically implemented robotic system that applies techniques from artificial intelligence to carry out cycles of scientific experimentation. The system automatically originates hypotheses to explain observations, devises experiments to test these hypotheses, physically runs the experiments using a laboratory robot, interprets the results to falsify hypotheses inconsistent with the data, and then repeats the cycle. Here we apply the system to the determination of gene function using deletion mutants of yeast (Saccharomyces cerevisiae) and auxotrophic growth experiments. We built and tested a detailed logical model (involving genes, proteins and metabolites) of the aromatic amino acid synthesis pathway. In biological experiments that automatically reconstruct parts of this model, we show that an intelligent experiment selection strategy is competitive with human performance and significantly outperforms, with a cost decrease of 3-fold and 100-fold (respectively), both cheapest and random-experiment selection.
New Scientist also had an article on it: "Robot scientist outperforms humans in lab."
I'm fairly certain that Ray Kurzweil and Hans Moravec at least know of each other, having served together on various panels. I wouldn't be surprised if the charts were used with permission. Considering that the overlap of their readership is pretty large, blatant plagiarism would be pretty easy to catch.
Deja vu ;)
Seriously though, this would be a very good idea. If such a prize existed, I'd certainly be willing to toss in some cash.
One of the bigest problem with neural networks is that 99.99% of all implementations are linear.
Do you have something to cite for that statement? I'm under the impression that just about everybody doing things with neural networks incorporates at least some sort of nonlinearity. Sigmoid functions and radial basis functions seem most popular nowadays, with the occasional spiking neural network thrown in to spice things up.
My personal opinion is that there was way too much of an emphasis on supervised learning in the 80s and 90s -- basically just finding ways to match up inputs and outputs. IMHO, unsupervised learning is where the fun stuff is.
This reminds me a little bit of a rather neat system I came across the other day, Video Google (despite the name, I don't think it has anything to do with the Google company). It doesn't use metadata or cheats like that, but rather uses image analysis to identify recurring objects and scenes.
They have a demo on their web site where you can select a portion of a video frame, and it'll show you all the places in the movie where the algorithm thinks that snippet shows up. Some other cool examples are displaying the appearances of a clock from 'Groundhog Day," and a recurring poster from 'Run Lola Run.' A research paper with more details is available here.
The abstract:
We describe an approach to object and scene retrieval which searches for and localizes all the occurrences of a user outlined object in a video. The object is represented by a set of viewpoint invariant region descriptors so that recognition can proceed successfully despite changes in view-point, illumination and partial occlusion. The temporal continuity of the video within a shot is used to track the regions in order to reject unstable regions and reduce the effects of noise in the descriptors.
The analogy with text retrieval is in the implementation where matches on descriptors are pre-computed (using vector quantization), and inverted file systems and document rankings are used. The result is that retrieval is immediate, returning a ranked list of key frames/shots in the manner of Google.
The method is illustrated for matching on two full length feature films.
Hating Bush is all well and good, but could you elaborate more on your space robotics -> militarization connection? You seem to be stating it rather axiomatically, and I'm afraid I don't follow your logic.
Disclaimer: I'm an avid roboticist and have done robotics research in the past. I'm fairly neutral on militarizing space, but am a very big advocate of developing robotic technologies, particularly because they're damned useful for pretty much any space-related activity.
(b) Bush cuts NASA spending (HST)
Not to nit-pick, but Bush has actually increased NASA spending. It's one of the few non-defense portions of the government which have actually gotten increased funding under his watch.
I already use programs like TiBR to read textfile novels on my Treo 600. It's a little scrunched, but I don't really notice it; things are probably considerably better on the Treo 650.
Project Gutenberg has plenty of textfile novels ripe for reading. Cory Doctorow's stuff is also pretty good. I read his Down and Out in the Magic Kingdom entirely on my cell.
Another handy resource is the University of Virginia Library's Etext Center, which has 1800 freely-available eBooks.
I think a better option than repair would be putting together the spares on the ground, including a PERFECT mirror as opposed to the flawed one that got launched, and launch Hubble II.
0
I agree that this would be good, although I doubt very much that it can be done by cobbling together leftovers. Last I saw, the estimated cost of such a thing was around $4B, vs. under $2B to fix the one we've got.
Actually, the cost of a new and much improved Hubble (including launch) would be more around $1B:
http://www.pha.jhu.edu/hop/
http://www.spaceref.com/news/viewpr.html?pid=1605
Nobody ever seems to mention the Hubble Origins Probe in these discussions, which is IMHO the best possible solution:
An international team led by Johns Hopkins University astronomers have proposed an alternative to sending a robotic or manned repair mission to the ailing Hubble Space Telescope. Their proposal is to build a new Hubble Origins Probe, reusing the Hubble design but using lighter and more cost-effective technologies. The probe would include instruments currently waiting to be installed on Hubble, as well as a Japanese-built imager which 'will allow scientists to map the heavens more than 20 times faster than even a refurbished Hubble Space Telescope could.' It would take an estimated 65 months and $1 billion to build and launch, approximately the same cost as a robotic service mission.
The original Hubble is great for historical and sentimental reasons, but the cost/benefit ratio is really so much better with a replacement.
Coincidentally, it was also posted on Slashdot on April 1, when it mentioned a PhysOrg story posted on March 29.
NASA should really start pushing for more private groups to do this rather than just handing out paltry prizes.
Unfortunately, merit-based funding of private projects and pork-barrel politics are rather at odds with each other.
Another point: Would you want to add an additional condition that all awardees would have to release the source code to their work, to allow future contestants to build on it? I believe the RoboCup robotic soccer competition does something like this.
The only problem with it is that it doesn't include Mahoney's threshold of 1.3 bits per character for artificial intelligence.
Indeed, but I don't see this as much of a problem. It favors steady incremental progress, which should eventually surpass the threshold.
Hehe. In any case, I think my Karma's been maxed out for a years now. If it helps my meta-karma, though, I've also contributed to some of the articles which I linked to.