We live in an accelerating culture. The difference between the change from 1901 to 2001 is far greater than the change from 1720 to 1820. If you imagine how confused someone living in during the turn of the (19 to 20th) century would feel if suddenly transported to today, you can be sure that it would be much more confusing to move somone from today to say.. 2101.
And if you don't believe that technology is accelerating, you might want to check out the number of patents issued in the past ten years versus the number of patents issued over say.. the 1960's. Even if you discount the fact that patents are arguably more often frivolous now, it still remains that invention is occuring on a much broader scale, at a faster rate.
A fundamental technology (like the electric motor, or the internet) provides a tremendous base for innovation. It seems to take a good number of years before the technology is sort of "taken for granted". There are many potential applications and adaptations of the Internet and its fundamental technologies that we haven't even begun to explore (and no, I don't know what possibilites are, else I would be on a path to big $$$).
Its also interesting to note, that the time for society to accept new technology on a widespread basis has also decreased. Example: Television took between 20 - 30 years before it could be considered a widespread technology. VCR's took 10 - 15 years. CD's 8 - 10. DVD's, while not quite widespread, are looking like they are going to be around 5 years or so (from the time that a large number of commericial players were offered to consumers, to their (relatively) widespread acceptance, and usage). It seems to me that it took people alot less time to get used to dvd's than cd's (I judge this from working at a library which didn't have cd's in its collection until the early 90's, but started offering dvd's in 1999).
So, change is not necessarily change, when you consider the fact that the amount of change taking place in a given period of time is increasing at a dramatic rate.
We are required to have a laptop by the end of our first year. The requirement specs at the beginning of last [school] year were for a pent3 500mhz machine w/ 802.11b card, running about $2k-2.5k. Granted costs have come down, but its still quite a bit for something that is required. The offical policy is that they want everyone to have laptops so they don't use the computer laps for "personal computing", aka typing reports, surfing net, email.. Makes you wonder what all of the labs full of PC's are for then.
In any case, I guess its cool that you can sit outside on the commons and get internet access.
Although, I'm almost tempted to pay a few bucks for a good FTP server capable of file ratios, bandwidth limits, and a built-in scheduler to disable/enable the server.
War ftp server. Windows 95/98/nt/2000, free, supports all the features you mentioned.
I'm not sure if its really neural net, as it is evolutive construction. The circuit in the article essentially evolved, through the natural selection that the researcher implemented, to a final state that accomplished the assigned task using fewer parts than anyone thought possible, and by using an inherent part of itself: feedback.
While there was not really any intelligence demonstrated in this particular example, there is certainly something here. The idea of letting the machine construct itself to meet an end goal, and not worry about implementation has huge implications for CS. Imagine a design specification language, in which a programmer writes detailed specs for a program, feeds it to an evolutive compiler, and out pops a program.
Join together an evolutive compiler, and maybe a neural net, and you have a program that learns and refines itself, using its own criteria of selection, and after enough iterations, perhaps you have a semi-intelligent program/thing.
I suppose it stands to reason that if it took nature trillions upon trillions of of organisms to create intelligent life, our attempts at artificial intelligence, at least along an evolutive path, would take roughly the same number of iterations before success. Here is the current hardware limitation.
...... Of course, by the same reasoning, if you put 10,000 monkeys in a room with 10,000 typewriters. . . .
I don't think the point of the device is to make playing the games any more effective. You'll never be able to kick thru a beam faster than you can mash a button with your thumb.
But imagine if game developers made parts of their games require a device like this. Instead of using it as an optional controller, it would be necessary to complete parts of the game using it. The game could detect which controller you were using, and perhaps only let you play certain levels using the physical input controller. It certainly would make alot of games quite a bit harder.
Slashdot Mirror
We live in an accelerating culture. The difference between the change from 1901 to 2001 is far greater than the change from 1720 to 1820. If you imagine how confused someone living in during the turn of the (19 to 20th) century would feel if suddenly transported to today, you can be sure that it would be much more confusing to move somone from today to say.. 2101.
And if you don't believe that technology is accelerating, you might want to check out the number of patents issued in the past ten years versus the number of patents issued over say.. the 1960's. Even if you discount the fact that patents are arguably more often frivolous now, it still remains that invention is occuring on a much broader scale, at a faster rate.
A fundamental technology (like the electric motor, or the internet) provides a tremendous base for innovation. It seems to take a good number of years before the technology is sort of "taken for granted". There are many potential applications and adaptations of the Internet and its fundamental technologies that we haven't even begun to explore (and no, I don't know what possibilites are, else I would be on a path to big $$$).
Its also interesting to note, that the time for society to accept new technology on a widespread basis has also decreased. Example: Television took between 20 - 30 years before it could be considered a widespread technology. VCR's took 10 - 15 years. CD's 8 - 10. DVD's, while not quite widespread, are looking like they are going to be around 5 years or so (from the time that a large number of commericial players were offered to consumers, to their (relatively) widespread acceptance, and usage). It seems to me that it took people alot less time to get used to dvd's than cd's (I judge this from working at a library which didn't have cd's in its collection until the early 90's, but started offering dvd's in 1999).
So, change is not necessarily change, when you consider the fact that the amount of change taking place in a given period of time is increasing at a dramatic rate.
We are required to have a laptop by the end of our first year. The requirement specs at the beginning of last [school] year were for a pent3 500mhz machine w/ 802.11b card, running about $2k-2.5k. Granted costs have come down, but its still quite a bit for something that is required. The offical policy is that they want everyone to have laptops so they don't use the computer laps for "personal computing", aka typing reports, surfing net, email.. Makes you wonder what all of the labs full of PC's are for then.
In any case, I guess its cool that you can sit outside on the commons and get internet access.
Although, I'm almost tempted to pay a few bucks for a good FTP server capable of file ratios, bandwidth limits, and a built-in scheduler to disable/enable the server.
War ftp server. Windows 95/98/nt/2000, free, supports all the features you mentioned.
I'm not sure if its really neural net, as it is evolutive construction. The circuit in the article essentially evolved, through the natural selection that the researcher implemented, to a final state that accomplished the assigned task using fewer parts than anyone thought possible, and by using an inherent part of itself: feedback.
...... Of course, by the same reasoning, if you put 10,000 monkeys in a room with 10,000 typewriters. . . .
While there was not really any intelligence demonstrated in this particular example, there is certainly something here. The idea of letting the machine construct itself to meet an end goal, and not worry about implementation has huge implications for CS. Imagine a design specification language, in which a programmer writes detailed specs for a program, feeds it to an evolutive compiler, and out pops a program.
Join together an evolutive compiler, and maybe a neural net, and you have a program that learns and refines itself, using its own criteria of selection, and after enough iterations, perhaps you have a semi-intelligent program/thing.
I suppose it stands to reason that if it took nature trillions upon trillions of of organisms to create intelligent life, our attempts at artificial intelligence, at least along an evolutive path, would take roughly the same number of iterations before success. Here is the current hardware limitation.
I guess I'm arguing that intelligence is a function of pathway complexity and self-referentiality (real word?).
Its being done, at least sort of. (Check here if you don't have access to New Scientist)
I wouldn't be surprised if a large portion of hardware and software is created this way in the future.
The converse is can the calculator understand algebra or calculus?
TI-89. I never would have passed calculus without it.
I don't think the point of the device is to make playing the games any more effective. You'll never be able to kick thru a beam faster than you can mash a button with your thumb. But imagine if game developers made parts of their games require a device like this. Instead of using it as an optional controller, it would be necessary to complete parts of the game using it. The game could detect which controller you were using, and perhaps only let you play certain levels using the physical input controller. It certainly would make alot of games quite a bit harder.