... I doubt that these are the needles they [the NSA] are seeking.
Yeah, but why not? This sort of thing obviously undercuts national security by tying up our cops, making them complicit in some asshole's prank, and causing potentially deadly danger. And compared to the effort and expense of mobilizing and deploying a freaking SWAT team, it is a comparative trifle for the NSA to answer a call from the cops asking for the malicious report to be traced to its source.
This is one of those times when our government's all knowing, all access panopticon would actually be useful. Seriously, our cops can't trace the swatters?
Facial recognition and object recognition was always thought of in the AI community as a "pattern interpretation" skill, and we suspected that human brains have special magic gears for "effortlessly" succeeding at these sorts of tasks, while AI coders struggled to emulate our success.
Now we're seriously talking about computers already being better at these tasks than we are. This is one of those milestones in AI research when we have to cross off another item from the list of "things that keep AIs from matching or exceeding human intelligence". For now, there are still many items on that list, but I wonder which ones will be crossed of next, and how soon.
When Assange went to hide in the Ecuadoran embassy, he had the world's attention. But now looking back, it looks like he basically put himself in prison, and not in the heroic way. If the US had grabbed him, tried him in some kangaroo court and imprisoned him, he'd stay relevant as a sort of journalistic martyr. But his embassy self-imprisonment left him just as isolated, while also looking vaguely like a fugitive instead of a martyr. Basically, he's been rendered irrelevant, without anything actually being done to him, apart from some sketchy accusation from two Swedish women. If this was some political chess match, Assange lost to a far superior opponent. I wouldn't be surprised if they just let him walk now. Either way Wikileaks has been killed without its killers having done anything that looks like a heavy-handed suppression of journalism.
For some reason, when men buy fancy Swiss watches these days, they buy these fat monstrosities that don't play well with the shirt cuff. The Apple watch is fat, but not fatter than many of the dumb watches from Switzerland. There are also some wonderful thin Swiss mechanical watches, but they aren't in fashion right now. If the Swiss push a new generation of sturdy but slim, elegant mechanical watches, I think they will be able to convince many men that they are better off choosing a wristwatch over some gaudy thing that duplicates the functions of their phone.
Their model basically assumes that you - the person who read the study - would be the only one who would flee to some location where you expect to be safe, and everyone else would stay where they were. If humans really were like that then by all means, follow the advice. But of course, many other humans would react to a zombie apocalypse by fleeing to the country. Quite probably, some would bring infected (still asymptomatic) victims along, which would infect others in the "isolated" sanctuary. How many residents from LA would drive to Death Valley because it seems like a place where zombies wouldn't be? Well guess what: That immigration wave is exactly how zombies get there. A better model would account for this predictable human flight behavior before arriving at a final recommendation.
I think we'd feel much better about it if we used this tool to remove suspicion from people, rather than to add suspicion. For example, this tool could probably rule out that the suspect is black or asian, that it's not a woman, it's not someone over 5'9, etc. Using the tool to generate a crappy portrait is the real bad move, because if you look like that, people will think that's evidence for your guilt. If this tool were only used to exonerate people and to remove them from the suspect list, who could object?
This might be useful if we ever build very slow, small and cheap interstellar colonization ships. Basically, I'm picturing something like a seed from which an entire civilization could hatch. In practice, it would be a tiny fabrication plant, plus lots of data. Once it arrives, the thing would use material from an asteroid or a comet to build larger and more specialized 3D printers, which would turn asteroids into a habitable space station, bioreplicators, etc. The bioreplicators would produce living germ cells from DNA data, artificial wombs would gestate them, and very fancy AI would parent the kids that come out. It's fun to think about how tiny the initial payload could be so that it's still big enough to eventually get the job done. Probably, the best way to do it would be to start with a single crude and tiny 3D printer, which is able to make a larger, better 3D printer, and so on.
Obviously, a big proportion of the mass of this thing would be the storage medium that carries all the data, because you won't just need software, videos, libraries, etc. You'll also need genetic info for an adequately diverse population of humans, plus an adequately diverse population of all the other living things those humans will need and want to have around, like gut bacteria, broccoli, earthworms, butterflies, kitties, etc. That's a lot of data, so you obviously want a robust and low-mass storage medium for it. The trip might take thousands of years, and space can be nasty. But if this DNA-in-glass medium can reliably last millions of years - more at 3 degrees K, I presume - maybe it would do the job. It would be really cool if it turned out to be possible to reconstitute our civilization in another solar system from a seed no larger than a trashcan. I don't see any reason to think it's impossible to go even smaller, maybe to the size of a beer can. The smaller it is, the easier it is to accelerate and decelerate. If it rides a laser beam on the way out, and decelerates with solar sail (like a parachute), it might be practical to make thousands of these rather cheaply. Any seeds that germinate could then make thousands more. (I think Freeman Dyson once discussed an idea like this.)
I know this sounds terribly traditional, but what could be wrong with sending a friend a letter in which you give instructions to post an update to social media on your behalf? I'm sure that all letters from prison would be read to make sure they're not carrying out something illegal, but it's not illegal for the friend to post an online update, right?
Or how about this: The friend starts a blog called "Letters From Sam in Jail" and just posts a scan of each letter received. That's a clear case where the prisoner is (indirectly) blogging, but nobody is doing something wrong. Right?
The point is that a computer probably generated a whole lot of poetry, and some poor human had to sift through it and pick the least awful poem. So, really, it's a human who did all the hard work anyway. You give enough monkeys enough typewriters, and hire some humans to sort through their "work" and you will eventually get something interesting too.
If Google cabs come pick you up and you pay them to drive you somewhere, Google is running a straight up taxi service. It's not ridesharing in any sense. Maybe Google would allow private car owners to put their driverless cars into the system, and keep a portion of the fares, but I don't see this as being very motivated. Google will have the driverless cars first, private competitors in their system would only drive down prices, and then there's the legwork of making sure that all the privateer taxis are safe and insured.
I love the idea of driverless taxis, and I'd love to live in a city where they were the only passenger cars allowed on roads. Unfortunately, I think that idiots will ruin the idea - for example, by using these things as convenient "date rape cabins".
I'm sure this happens already, as kids often add absurd fins and wings to production cars, and then drive the modified abominations on public roads. But apparently I've been OK with this for my whole life, so I have no reason to worry about what will essentially be more of the same.
I notice that it's been a while since I've worried about document formats. I'm not so vain as to need features not supported in Rich Text Format, so for 20 years I've been sending people.rtf files out of compatibility politeness. Once, when I explained all this to someone, the response I got was something like "Dude, these days, everyone can open basically everything." And there's something right about that. In the old days I worried about formats forcing MS Office lock-in, but nowadays it's hard to get me too worked up about it. It's kind of like video files, where you barely notice whether you downloaded a.mpg,.avi,.wmv, mkv, or mp4. You double-click and it all plays.
I'm sure the lowest-spec version with the four extra gig-E ports would be more than adequate. Consumer routers sometimes cost more, but they don't have SATA and their USB throughput suuucks.
Your post is suggesting that Musk may be naive enough to not anticipate all the difficult steps necessary for ramping up production, but somehow I have more faith in him as a businessman. Did you see how he handled the negotiations about the battery factory? Like a bawss.
Higher end Teslas were always meant to be a learning experience for the company. They're high-end cars, so they're made in a boutique setting like high end cars tend to be. There is a sense in starting upmarket and working your way down as you find your feet. You can bet that Musk has very smart people identifying and fixing all production bottlenecks. When they're ironed out, the cost of production really will fall a lot, because the sense behind his process is to minimize human intervention. The problem with automation is designing machines that are good enough, but I expect that this is exactly where they're making headway. Well, maybe not. And maybe, demand for Teslas isn't bottomless, once they're produced on a BMW scale. But I wouldn't bet against Musk.
Maybe it's the season, but doesn't this sound like like a bunch of overindulged, adult children in uniforms, sitting around a table trying to figure out what toys they don't yet have, which might be fun to play with? Like, they're so bored with quadcopters now, they want a fucking hawk. Because fuck yeah, hawk. Taxpayers should buy them a mechanical hawk.
I really fucking hate this about academia. It's absolutely shameless to charge college students $244 for a single dumb textbook. It's not even that good. It's just that when a department chooses to standardize on a textbook, the move has inertia and is basically impossible to reverse. Then, the publisher can charge something absurd, and everybody pays it, because it is a required text. It's so dirty, because it's profiteering from people who are often barely making ends meet, and typically buying the book with debt.
What really bothers me is that nobody seems willing to do anything about it. If a big, publicly funded university system set aside some money to create and regularly update their core STEM curriculum textbooks - let's start with Calculus, Physics, GenChem, GenBio - it would certainly cost less than the almost $1000 per student that the textbook purchases cost. These universities have Nobel Prize winners among their faculty, surely they have the in-house resources to create excellent textbooks and distribute them on some sort of open license like CC. Arranging sabbaticals for the authors might cost at most a million dollars, or roughly 4000 Stewart Calculus books. That might be about the number of Calc 1, Phys 1, GenChem and GenBio books that are sold on a single campus in a single year.
But this move would help everybody, not just within the entire UC system that funded the effort, but across the globe. And the costs of updating and embellishing future editions would be far less. I'm so mad that a large university system doesn't just make this happen. And yes, raise fucking tuition by $200 to pay for it, if you absolutely have to. In exchange for textbooks you can have for free (or for printing cost if you don't like digital), everybody will recognize that's a great deal. The courses can explicitly invite students to devise problems for future editions, or to suggest changes and clarifications. And it will bring prestige to the colleges and to the authors, which is worth something too.
Requests for ACH transfers are collected by banks and submitted in batches, once a day, and the banks receiving the transfers also process the payments once a day, leading to long waits. ACH technology was created in the 1970s and has not changed significantly since.
Yeah, except the Spanish media is not at all in a good negotiating position. It's not like the only Spanish-language press is in Spain. Spaniards who like Google's service can just switch their link to news.google.ar,.mx, or whatever. Or Google can even keep news.google.es but focus on stories about Spain as they appear in the Spanish-speaking press outside of Spain.
If Spaniards come to see domestic newspapers as dispensable, those newspapers are the only party that loses. In fact, I would bet that before long, some of the minor Spanish news outlets will break and announce that they have arranged an fee exemption for Google news. Without domestic competition, these sources will suddenly have top billing and a surge in traffic. And suddenly, everyone else will announce their own fee exemption, and this whole thing will end how it started.
I hope that these videolectures.net versions are not the only ones that will be saved. They look like re-encodes of the already ugly 240p version. I know all the artifacts don't make it unintelligible, but they are very distracting, especially if you watch at full screen. At some point, MIT re-capped the videotape with much better capture hardware and in 720p. That's what you got to watch if you did the MIT-X course.
MIT's video was 720p - granted, their source material was VHS, but it looks far better than the re-encoded thing on videolectures.net. I found I have a local copy of the first seven lectures of 8.0.1 in 720p, and I'm sure that others have the rest. Especially because it's CC, I'd hate for these to just be lost.
I have about the first half of the 801 course (Newtonian mechanics) in 720p. I downloaded the videos so I could watch them during my commute when I was offline. The other videos must be on someone's hard drive somewhere, right? I think it's time for some sneaky guerrilla distributed archiving!
Slashdot Mirror
... I doubt that these are the needles they [the NSA] are seeking.
Yeah, but why not? This sort of thing obviously undercuts national security by tying up our cops, making them complicit in some asshole's prank, and causing potentially deadly danger. And compared to the effort and expense of mobilizing and deploying a freaking SWAT team, it is a comparative trifle for the NSA to answer a call from the cops asking for the malicious report to be traced to its source.
This is one of those times when our government's all knowing, all access panopticon would actually be useful. Seriously, our cops can't trace the swatters?
Facial recognition and object recognition was always thought of in the AI community as a "pattern interpretation" skill, and we suspected that human brains have special magic gears for "effortlessly" succeeding at these sorts of tasks, while AI coders struggled to emulate our success.
Now we're seriously talking about computers already being better at these tasks than we are. This is one of those milestones in AI research when we have to cross off another item from the list of "things that keep AIs from matching or exceeding human intelligence". For now, there are still many items on that list, but I wonder which ones will be crossed of next, and how soon.
If the nine jurors that rendered the verdict had to compose the firing squad, I would feel somewhat better about the death penalty.
When Assange went to hide in the Ecuadoran embassy, he had the world's attention. But now looking back, it looks like he basically put himself in prison, and not in the heroic way. If the US had grabbed him, tried him in some kangaroo court and imprisoned him, he'd stay relevant as a sort of journalistic martyr. But his embassy self-imprisonment left him just as isolated, while also looking vaguely like a fugitive instead of a martyr. Basically, he's been rendered irrelevant, without anything actually being done to him, apart from some sketchy accusation from two Swedish women. If this was some political chess match, Assange lost to a far superior opponent. I wouldn't be surprised if they just let him walk now. Either way Wikileaks has been killed without its killers having done anything that looks like a heavy-handed suppression of journalism.
... record labels are going to become more reluctant to release music that's similar to other works...
But... but... how will civilization survive?
For some reason, when men buy fancy Swiss watches these days, they buy these fat monstrosities that don't play well with the shirt cuff. The Apple watch is fat, but not fatter than many of the dumb watches from Switzerland. There are also some wonderful thin Swiss mechanical watches, but they aren't in fashion right now. If the Swiss push a new generation of sturdy but slim, elegant mechanical watches, I think they will be able to convince many men that they are better off choosing a wristwatch over some gaudy thing that duplicates the functions of their phone.
Their model basically assumes that you - the person who read the study - would be the only one who would flee to some location where you expect to be safe, and everyone else would stay where they were. If humans really were like that then by all means, follow the advice. But of course, many other humans would react to a zombie apocalypse by fleeing to the country. Quite probably, some would bring infected (still asymptomatic) victims along, which would infect others in the "isolated" sanctuary. How many residents from LA would drive to Death Valley because it seems like a place where zombies wouldn't be? Well guess what: That immigration wave is exactly how zombies get there. A better model would account for this predictable human flight behavior before arriving at a final recommendation.
I think we'd feel much better about it if we used this tool to remove suspicion from people, rather than to add suspicion. For example, this tool could probably rule out that the suspect is black or asian, that it's not a woman, it's not someone over 5'9, etc. Using the tool to generate a crappy portrait is the real bad move, because if you look like that, people will think that's evidence for your guilt. If this tool were only used to exonerate people and to remove them from the suspect list, who could object?
This might be useful if we ever build very slow, small and cheap interstellar colonization ships. Basically, I'm picturing something like a seed from which an entire civilization could hatch. In practice, it would be a tiny fabrication plant, plus lots of data. Once it arrives, the thing would use material from an asteroid or a comet to build larger and more specialized 3D printers, which would turn asteroids into a habitable space station, bioreplicators, etc. The bioreplicators would produce living germ cells from DNA data, artificial wombs would gestate them, and very fancy AI would parent the kids that come out. It's fun to think about how tiny the initial payload could be so that it's still big enough to eventually get the job done. Probably, the best way to do it would be to start with a single crude and tiny 3D printer, which is able to make a larger, better 3D printer, and so on.
Obviously, a big proportion of the mass of this thing would be the storage medium that carries all the data, because you won't just need software, videos, libraries, etc. You'll also need genetic info for an adequately diverse population of humans, plus an adequately diverse population of all the other living things those humans will need and want to have around, like gut bacteria, broccoli, earthworms, butterflies, kitties, etc. That's a lot of data, so you obviously want a robust and low-mass storage medium for it. The trip might take thousands of years, and space can be nasty. But if this DNA-in-glass medium can reliably last millions of years - more at 3 degrees K, I presume - maybe it would do the job. It would be really cool if it turned out to be possible to reconstitute our civilization in another solar system from a seed no larger than a trashcan. I don't see any reason to think it's impossible to go even smaller, maybe to the size of a beer can. The smaller it is, the easier it is to accelerate and decelerate. If it rides a laser beam on the way out, and decelerates with solar sail (like a parachute), it might be practical to make thousands of these rather cheaply. Any seeds that germinate could then make thousands more. (I think Freeman Dyson once discussed an idea like this.)
I know this sounds terribly traditional, but what could be wrong with sending a friend a letter in which you give instructions to post an update to social media on your behalf? I'm sure that all letters from prison would be read to make sure they're not carrying out something illegal, but it's not illegal for the friend to post an online update, right?
Or how about this: The friend starts a blog called "Letters From Sam in Jail" and just posts a scan of each letter received. That's a clear case where the prisoner is (indirectly) blogging, but nobody is doing something wrong. Right?
The point is that a computer probably generated a whole lot of poetry, and some poor human had to sift through it and pick the least awful poem. So, really, it's a human who did all the hard work anyway. You give enough monkeys enough typewriters, and hire some humans to sort through their "work" and you will eventually get something interesting too.
If Google cabs come pick you up and you pay them to drive you somewhere, Google is running a straight up taxi service. It's not ridesharing in any sense. Maybe Google would allow private car owners to put their driverless cars into the system, and keep a portion of the fares, but I don't see this as being very motivated. Google will have the driverless cars first, private competitors in their system would only drive down prices, and then there's the legwork of making sure that all the privateer taxis are safe and insured.
I love the idea of driverless taxis, and I'd love to live in a city where they were the only passenger cars allowed on roads. Unfortunately, I think that idiots will ruin the idea - for example, by using these things as convenient "date rape cabins".
I'm sure this happens already, as kids often add absurd fins and wings to production cars, and then drive the modified abominations on public roads. But apparently I've been OK with this for my whole life, so I have no reason to worry about what will essentially be more of the same.
The compatibility patch from Microsoft has been out since forever. Office 2003 reads and writes .docx files without any trouble.
I notice that it's been a while since I've worried about document formats. I'm not so vain as to need features not supported in Rich Text Format, so for 20 years I've been sending people .rtf files out of compatibility politeness. Once, when I explained all this to someone, the response I got was something like "Dude, these days, everyone can open basically everything." And there's something right about that. In the old days I worried about formats forcing MS Office lock-in, but nowadays it's hard to get me too worked up about it. It's kind of like video files, where you barely notice whether you downloaded a .mpg, .avi, .wmv, mkv, or mp4. You double-click and it all plays.
I'm sure the lowest-spec version with the four extra gig-E ports would be more than adequate. Consumer routers sometimes cost more, but they don't have SATA and their USB throughput suuucks.
Your post is suggesting that Musk may be naive enough to not anticipate all the difficult steps necessary for ramping up production, but somehow I have more faith in him as a businessman. Did you see how he handled the negotiations about the battery factory? Like a bawss.
Higher end Teslas were always meant to be a learning experience for the company. They're high-end cars, so they're made in a boutique setting like high end cars tend to be. There is a sense in starting upmarket and working your way down as you find your feet. You can bet that Musk has very smart people identifying and fixing all production bottlenecks. When they're ironed out, the cost of production really will fall a lot, because the sense behind his process is to minimize human intervention. The problem with automation is designing machines that are good enough, but I expect that this is exactly where they're making headway. Well, maybe not. And maybe, demand for Teslas isn't bottomless, once they're produced on a BMW scale. But I wouldn't bet against Musk.
Maybe it's the season, but doesn't this sound like like a bunch of overindulged, adult children in uniforms, sitting around a table trying to figure out what toys they don't yet have, which might be fun to play with? Like, they're so bored with quadcopters now, they want a fucking hawk. Because fuck yeah, hawk. Taxpayers should buy them a mechanical hawk.
I really fucking hate this about academia. It's absolutely shameless to charge college students $244 for a single dumb textbook. It's not even that good. It's just that when a department chooses to standardize on a textbook, the move has inertia and is basically impossible to reverse. Then, the publisher can charge something absurd, and everybody pays it, because it is a required text. It's so dirty, because it's profiteering from people who are often barely making ends meet, and typically buying the book with debt.
What really bothers me is that nobody seems willing to do anything about it. If a big, publicly funded university system set aside some money to create and regularly update their core STEM curriculum textbooks - let's start with Calculus, Physics, GenChem, GenBio - it would certainly cost less than the almost $1000 per student that the textbook purchases cost. These universities have Nobel Prize winners among their faculty, surely they have the in-house resources to create excellent textbooks and distribute them on some sort of open license like CC. Arranging sabbaticals for the authors might cost at most a million dollars, or roughly 4000 Stewart Calculus books. That might be about the number of Calc 1, Phys 1, GenChem and GenBio books that are sold on a single campus in a single year.
But this move would help everybody, not just within the entire UC system that funded the effort, but across the globe. And the costs of updating and embellishing future editions would be far less. I'm so mad that a large university system doesn't just make this happen. And yes, raise fucking tuition by $200 to pay for it, if you absolutely have to. In exchange for textbooks you can have for free (or for printing cost if you don't like digital), everybody will recognize that's a great deal. The courses can explicitly invite students to devise problems for future editions, or to suggest changes and clarifications. And it will bring prestige to the colleges and to the authors, which is worth something too.
Requests for ACH transfers are collected by banks and submitted in batches, once a day, and the banks receiving the transfers also process the payments once a day, leading to long waits. ACH technology was created in the 1970s and has not changed significantly since.
Jesus Christ. How much do we pay these people?
Yeah, except the Spanish media is not at all in a good negotiating position. It's not like the only Spanish-language press is in Spain. Spaniards who like Google's service can just switch their link to news.google.ar, .mx, or whatever. Or Google can even keep news.google.es but focus on stories about Spain as they appear in the Spanish-speaking press outside of Spain.
If Spaniards come to see domestic newspapers as dispensable, those newspapers are the only party that loses. In fact, I would bet that before long, some of the minor Spanish news outlets will break and announce that they have arranged an fee exemption for Google news. Without domestic competition, these sources will suddenly have top billing and a surge in traffic. And suddenly, everyone else will announce their own fee exemption, and this whole thing will end how it started.
I hope that these videolectures.net versions are not the only ones that will be saved. They look like re-encodes of the already ugly 240p version. I know all the artifacts don't make it unintelligible, but they are very distracting, especially if you watch at full screen. At some point, MIT re-capped the videotape with much better capture hardware and in 720p. That's what you got to watch if you did the MIT-X course.
MIT's video was 720p - granted, their source material was VHS, but it looks far better than the re-encoded thing on videolectures.net. I found I have a local copy of the first seven lectures of 8.0.1 in 720p, and I'm sure that others have the rest. Especially because it's CC, I'd hate for these to just be lost.
I have about the first half of the 801 course (Newtonian mechanics) in 720p. I downloaded the videos so I could watch them during my commute when I was offline. The other videos must be on someone's hard drive somewhere, right? I think it's time for some sneaky guerrilla distributed archiving!