True, but I think the more likely spoof would be something like "bankófamerica.com". If the history of phishing is a guide, that will net some users. I don't know if they are already considering such a thing, but limiting each component in the URL to a single language character set might be a good idea.
To me, the best approach would be to limit the character sets that can appear below any given TLD. So, having simplified Chinese under ".ch" would be fine, but not under ".com" or ".us" -- The idea being that local users should be able to tell the difference. That's because the biggest problem, as I see it, lies when people get exposed to characters they aren't used to; Our brains tend to map unfamiliar symbols to the ones we know.
If you don't want to mess with detailed setup, why the hell are you running Slackware? That's like buying a car with a manual transmission and then complaining that you have to switch gears...
While you're right, I'm not sure that people really consider traditional AI to be dead. Well, all I claimed was that the *hype* is dead, which I think most people will agree on. Nothing in research ever truly "dies" (unless everyone stops working on it and all the papers are lost); Instead it merely slows down depending on how few people are working on it.
Certainly it's been said that once we know how to do it, then people stop calling it AI. Yeah I certainly didn't invent the quote, but I don't know where it originated. Hearing it at AAAI would certainly make sense. My advisor's background was in symbolic planning, so she certainly has connections with many of those researchers. I'm more of a robotics/algorithms person myself.
If I were to draw a distinction between "old AI" and "new AI", I'd probably say that old AI was driven primarily by logic and representation, while "new AI" is primarily driven by data and inductive inference. Old AI is still grappling with the problem that representation, which is really really difficult, while new AI is trying to build up from what we currently have, which is sensors that give us lots of data. The new AI people (perhaps naively) hope that we can build up steadily toward intelligence, and if we're lucky, find a way to make an end run around the difficulty of explicit representation. That's the idea at least, but I am happy that people are attacking the problem from all angles. For one example, I've been quote impressed with the last 10 years of research in symbolic planning -- if we can find a suitable representation, I'm confident we'll have the planners to work with it.
Thank you; I was going to make a similar point. In short, evolution created intelligent beings, yet evolution does not itself understand consciousness. Intelligence did take a long time to arise, but with guided evolution (i.e. research and design), machines will get there much faster. Hans Moravec likes to point out that computing power is evolving 5 million times faster than in nature. Even if robots, algorithms, and AI lag by a significant factor, we're still likely to have something useful within our lifetimes, and something operating at a near-human level by the next century.
While much of the "traditional AI" hype could be considered dead, robotics is continuing to advance, and much symbolic AI research has evolved into data-driven statistical techniques. So while the top-down ideas that the older AI researches didn't pan out yet, bottom-up techniques will still help close the gap.
Also, you have to remember that AI is pretty much defined as "the stuff we don't know how to do yet". Once we know how to do it, then people stop calling it AI, and then wonder "why can't we do AI?" Machine vision is doing everything from factory inspections to face recognition, we have voice recognition on our cell phones, and context-sensitive web search is common. All those things were considered AI not long ago. Calculators were once even called mechanical brains.
Yes. He founded the Leg Lab at Carnegie Mellon, and later took it to MIT where it got even more famous. Then he left for Boston Dynamics, which created the "big-dog" robot that has been on slashdot. Gill Pratt continued the MIT Leg Lab work on force-based actuation and dynamic walking for some time after that.
The F22 does not normally have external mounts, but there are hardpoints where they can be added. Of course that would defeat the stealth, but if you're about to drop two nukes, at that point you're probably beyond being sneaky.
Quick sort does not require random access in order to be E[O(n*log(n))] on a shuffled linked list. Here is a good example. Even if you want to use the middle element as a pivot (good for near-sorted data), you can scan the linked list linearly for that element, and it will amortize to the same complexity.
Now, there is the question of whether you want to do it, but you most certainly can. For my unsigned integer linked-list sorting needs, I've found radix sort with a radix of 6 to work pretty well in practice.
Can't a prominent OSS person just switch anymore? It seems like they have to make a big political stink out of it. It's really too bad that people can't leave when things are still amicable, and instead they let it boil over to a traditional email flame-fest by the time they act.
It amazes me how much people have picked up Field D*, given that it drops the optimality associated with A*/D*. Field D* tends to return shorter paths, but it is not guaranteed to do so. Since those were the exact reasons that groups like NASA would not touch randomized planners, I find that a little odd. I guess its the nice even upgrade path of A*->D*->Field D*. At any rate, having NASA pick up any recent work on autonomy is a success, given their conservative nature.
I remember an Air Force person once saying they would never fund any research using any randomized algorithms; The funny thing is I managed to make a version of the randomized RRT planner the primary route planner for a UAV research project. Grids simply break down as soon as you add any sort of additional dimensions to the problem, and randomized approaches are perfect for 3D worlds and/or kinematic constraints. Personally I am a bit biased though, as I am a big proponent or randomized RRT/PRM based methods, even in low dimensions.
Now if only game programmers would notice that there's been some advances since A*...
Look at my robots and tell me they are dumb. NASA is not state of the art, they are very conservative. That's not always a bad thing when your robot is worth many millions of dollars and is halfway across the solar system. However, don't think that people aren't doing research now for NASA to use in 2020.
Ugh, as a robotics person, I find that a bit insulting. NASA is overly conservative, but most game programmers don't even know there are other algorithms than A*. Their books have A*, so that's all you ever see. Most navigation graphs in 3D games aren't even automatically generated. No games use D*, field D*, or randomized roadmaps. Just 1970-80's era hierarchical A*. They'll use something from last year's Siggraph, but its like the last 15 years of robotics research never even happened.
I hope others read your post, as you nicely outlined the right way for most people to switch. Don't dive in, expecting everything to work immediately, and then jump out after you get burned; experiment. Your post made me realize that's how I actually switched; It was three years between the first time I tried installing Linux to when I stopped regularly booting Windows. In between, I tried installing a couple of distributions, and got familiar with Linux/Unix while using clusters at school. Good luck in your endeavors.
Frankly, I think CBS has the right idea with youtube; Short clips are a great way to advertise your TV shows, and also people will sit through 5-10 seconds ads in order to get the known quality of a specific uploader (i.e. CBS in this case). Right now, the only thing hurting them is that you can't do a youtube search limited to a specific user, and any random Joe can put "CBS" in his tags. I would expect google to fix that eventually.
Others will jump on the bandwagon when they realize CBS is benefiting from youtube. Joost is actually orthogonal, since it is about full-length shows rather than clips. Smart TV channels would likely post on both, using youtube clips to advertise shows on TV as well as Joost. Of course with the inevitable cracking of Joost's DRM, that will likely cause the channels to pull any new content, and doom joost's entire business model.
I dunno, if I were an alien race, and had the choice of colonizing solar systems where the best planet was: (1) Teeming with constrantly mutating alien bacteria (2) Lifeless and ready for terraforming
I know which one I would choose. Seriously, why risk alien disease when there are so many "clean" places to choose from? If you were looking for a cave to sleep in, would you choose the empty one or the one with animals already in it? Unless space travel is instant, I really don't see a race ever expanding fast enough to need to use every planet. Besides, it is selfish to think alien life is "as we know it" and would even care about our planet; If they aren't water-based our planet could seem like the same kind of hell that Venus seems to us.
As does clicking on a checkbox. Seriously, I don't need it to be any easier.
What we do need now are better guides telling people what they need to install for a given problem, since the link from task to program/package name is not always obvious. I know I can use google for that, but a more centralized and guided way would be nice for less technical people. That's why I think that click-n-run has its place.
Don't worry, a seventh way will come along to wrap those first six which don't solve the problem, and it will be the ultimate meta-universal generic packaging system.
No, it's more like not giving money to a homeless person who has a drinking problem, or not donating to your local sheriff's organization a day after the paper reports on widespread corruption in the department. Sometimes it works to let them know "if you clean up your act, I could help you." My money is finite, and I do have the choice of where I spend it, so I choose to spend it on organizations whose work is important, and where I think the money will be efficiently put to good use.
Slashdot Mirror
True, but I think the more likely spoof would be something like "bankófamerica.com". If the history of phishing is a guide, that will net some users. I don't know if they are already considering such a thing, but limiting each component in the URL to a single language character set might be a good idea.
To me, the best approach would be to limit the character sets that can appear below any given TLD. So, having simplified Chinese under ".ch" would be fine, but not under ".com" or ".us" -- The idea being that local users should be able to tell the difference. That's because the biggest problem, as I see it, lies when people get exposed to characters they aren't used to; Our brains tend to map unfamiliar symbols to the ones we know.
I heard that dead people prefer genuine Microsoft Windows(tm) to Linux by almost 4:1.
Sure, but when you've got A and B, why not patent AB!
Yeah, but with Windows, through all of that you never once needed to edit a scary config file. Wasn't it so much easier that way?
If you don't want to mess with detailed setup, why the hell are you running Slackware? That's like buying a car with a manual transmission and then complaining that you have to switch gears...
Netcraft may or may not confirm it, since it rarely can tell Linux versions or distros apart anyway.
No, but it's a hell of a lot better than asking Slashdot.
You forgot:
LISP programmers didn't even realize there was a difference.
If I were to draw a distinction between "old AI" and "new AI", I'd probably say that old AI was driven primarily by logic and representation, while "new AI" is primarily driven by data and inductive inference. Old AI is still grappling with the problem that representation, which is really really difficult, while new AI is trying to build up from what we currently have, which is sensors that give us lots of data. The new AI people (perhaps naively) hope that we can build up steadily toward intelligence, and if we're lucky, find a way to make an end run around the difficulty of explicit representation. That's the idea at least, but I am happy that people are attacking the problem from all angles. For one example, I've been quote impressed with the last 10 years of research in symbolic planning -- if we can find a suitable representation, I'm confident we'll have the planners to work with it.
Thank you; I was going to make a similar point. In short, evolution created intelligent beings, yet evolution does not itself understand consciousness. Intelligence did take a long time to arise, but with guided evolution (i.e. research and design), machines will get there much faster. Hans Moravec likes to point out that computing power is evolving 5 million times faster than in nature. Even if robots, algorithms, and AI lag by a significant factor, we're still likely to have something useful within our lifetimes, and something operating at a near-human level by the next century.
While much of the "traditional AI" hype could be considered dead, robotics is continuing to advance, and much symbolic AI research has evolved into data-driven statistical techniques. So while the top-down ideas that the older AI researches didn't pan out yet, bottom-up techniques will still help close the gap.
Also, you have to remember that AI is pretty much defined as "the stuff we don't know how to do yet". Once we know how to do it, then people stop calling it AI, and then wonder "why can't we do AI?" Machine vision is doing everything from factory inspections to face recognition, we have voice recognition on our cell phones, and context-sensitive web search is common. All those things were considered AI not long ago. Calculators were once even called mechanical brains.
Yes. He founded the Leg Lab at Carnegie Mellon, and later took it to MIT where it got even more famous. Then he left for Boston Dynamics, which created the "big-dog" robot that has been on slashdot. Gill Pratt continued the MIT Leg Lab work on force-based actuation and dynamic walking for some time after that.
The F22 does not normally have external mounts, but there are hardpoints where they can be added. Of course that would defeat the stealth, but if you're about to drop two nukes, at that point you're probably beyond being sneaky.
Quick sort does not require random access in order to be E[O(n*log(n))] on a shuffled linked list. Here is a good example. Even if you want to use the middle element as a pivot (good for near-sorted data), you can scan the linked list linearly for that element, and it will amortize to the same complexity.
Now, there is the question of whether you want to do it, but you most certainly can. For my unsigned integer linked-list sorting needs, I've found radix sort with a radix of 6 to work pretty well in practice.
Can't a prominent OSS person just switch anymore? It seems like they have to make a big political stink out of it. It's really too bad that people can't leave when things are still amicable, and instead they let it boil over to a traditional email flame-fest by the time they act.
It amazes me how much people have picked up Field D*, given that it drops the optimality associated with A*/D*. Field D* tends to return shorter paths, but it is not guaranteed to do so. Since those were the exact reasons that groups like NASA would not touch randomized planners, I find that a little odd. I guess its the nice even upgrade path of A*->D*->Field D*. At any rate, having NASA pick up any recent work on autonomy is a success, given their conservative nature.
I remember an Air Force person once saying they would never fund any research using any randomized algorithms; The funny thing is I managed to make a version of the randomized RRT planner the primary route planner for a UAV research project. Grids simply break down as soon as you add any sort of additional dimensions to the problem, and randomized approaches are perfect for 3D worlds and/or kinematic constraints. Personally I am a bit biased though, as I am a big proponent or randomized RRT/PRM based methods, even in low dimensions.
Now if only game programmers would notice that there's been some advances since A*...
Look at my robots and tell me they are dumb. NASA is not state of the art, they are very conservative. That's not always a bad thing when your robot is worth many millions of dollars and is halfway across the solar system. However, don't think that people aren't doing research now for NASA to use in 2020.
Ugh, as a robotics person, I find that a bit insulting. NASA is overly conservative, but most game programmers don't even know there are other algorithms than A*. Their books have A*, so that's all you ever see. Most navigation graphs in 3D games aren't even automatically generated. No games use D*, field D*, or randomized roadmaps. Just 1970-80's era hierarchical A*. They'll use something from last year's Siggraph, but its like the last 15 years of robotics research never even happened.
I hope others read your post, as you nicely outlined the right way for most people to switch. Don't dive in, expecting everything to work immediately, and then jump out after you get burned; experiment. Your post made me realize that's how I actually switched; It was three years between the first time I tried installing Linux to when I stopped regularly booting Windows. In between, I tried installing a couple of distributions, and got familiar with Linux/Unix while using clusters at school. Good luck in your endeavors.
ALL THESE WORLDS ARE YOURS EXCEPT EUROPA. ATTEMPT NO PATENT LITIGATION THERE.
the lameness filter is lame. the lameness filter is lame. the lameness filter is lame.
Here's a link backing it up.
Frankly, I think CBS has the right idea with youtube; Short clips are a great way to advertise your TV shows, and also people will sit through 5-10 seconds ads in order to get the known quality of a specific uploader (i.e. CBS in this case). Right now, the only thing hurting them is that you can't do a youtube search limited to a specific user, and any random Joe can put "CBS" in his tags. I would expect google to fix that eventually.
Others will jump on the bandwagon when they realize CBS is benefiting from youtube. Joost is actually orthogonal, since it is about full-length shows rather than clips. Smart TV channels would likely post on both, using youtube clips to advertise shows on TV as well as Joost. Of course with the inevitable cracking of Joost's DRM, that will likely cause the channels to pull any new content, and doom joost's entire business model.
I dunno, if I were an alien race, and had the choice of colonizing solar systems where the best planet was:
(1) Teeming with constrantly mutating alien bacteria
(2) Lifeless and ready for terraforming
I know which one I would choose. Seriously, why risk alien disease when there are so many "clean" places to choose from? If you were looking for a cave to sleep in, would you choose the empty one or the one with animals already in it? Unless space travel is instant, I really don't see a race ever expanding fast enough to need to use every planet. Besides, it is selfish to think alien life is "as we know it" and would even care about our planet; If they aren't water-based our planet could seem like the same kind of hell that Venus seems to us.
As does clicking on a checkbox. Seriously, I don't need it to be any easier.
What we do need now are better guides telling people what they need to install for a given problem, since the link from task to program/package name is not always obvious. I know I can use google for that, but a more centralized and guided way would be nice for less technical people. That's why I think that click-n-run has its place.
Don't worry, a seventh way will come along to wrap those first six which don't solve the problem, and it will be the ultimate meta-universal generic packaging system.
No, it's more like not giving money to a homeless person who has a drinking problem, or not donating to your local sheriff's organization a day after the paper reports on widespread corruption in the department. Sometimes it works to let them know "if you clean up your act, I could help you." My money is finite, and I do have the choice of where I spend it, so I choose to spend it on organizations whose work is important, and where I think the money will be efficiently put to good use.