To me this falls under the same category as "your right to swing your fists ends at my nose." That is , you have the right to express yourself (via spam or any other medium of your choice) as long as I have the right to ignore you.
Hmmm, I would say that the difference is that there are inherent barriers to the copying of that library book, namely that it's a lot of work to xerox the whole thing. If somebody invented a machine that could copy a book (nice binding and all) in under a minute, then the anology would be more solid.
Given that it's really easy to copy a digital work, it's reasonable to expect their owners to protect them, thereby creating a substitute legal barrier where there was no inherent barrier.
Well that analogy would be valid if I could walk away with a copy of your car and the original would still be there when you got back...
Re:Tracking encrypted communications
on
A New Kind of War
·
· Score: 1
This assumes that you can detect when a message is encrypted; if not (see my above post), you have to wade through millions of people who happen to be sending a lot more email than they are receiving.
Possibly desireable, but not necessary; techniques exist which allow embedded information to survive through compression, and even destruction of large chunks (half or more) of an image. Read up on "digital watermarking" -- I know Kodak has done a lot of research here.
Digital correspondence can always be buried in noise. A message could (for example) be written, heavily encrypted, and then embedded into a JPEG image in such a way that it would be undetectable without the right extraction software.
Does the government even have the processing power to do keyword searches on all the email that it monitors? Much less checking for the presence of heavily embedded messages...
I believe that any measures that the gov't takes to aid in electronic surveillence will only compromise our liberties.
I strongly agree with this. Anyone who has done research into artificial neural networks knows that the mathematics required to simulate them grows in complexity exponentially as the number of neurons in the network grow. We do not even posess the mathematical tools today to comprehensively model a network of ten neurons with a relatively simple activation function, unless we discretize the simulation in time.
I have done research which explored the presence of chaos in ANNs, and found that even such simple networks exhibit massive variances in outputs when very small changes are made to the parameters of their activation functions. And neuroscience is just scratching the surface of how a real neuron is activated electrochemically.
I think that the complexity of our own brains will push Kurzweil's predictions back at least a few centuries. But I'll save a copy of the interview for nostalgia's sake in 2050.
Slashdot Mirror
To me this falls under the same category as "your right to swing your fists ends at my nose." That is , you have the right to express yourself (via spam or any other medium of your choice) as long as I have the right to ignore you.
trb001: Show me the difference.
Hmmm, I would say that the difference is that there are inherent barriers to the copying of that library book, namely that it's a lot of work to xerox the whole thing. If somebody invented a machine that could copy a book (nice binding and all) in under a minute, then the anology would be more solid.
Given that it's really easy to copy a digital work, it's reasonable to expect their owners to protect them, thereby creating a substitute legal barrier where there was no inherent barrier.
Well that analogy would be valid if I could walk away with a copy of your car and the original would still be there when you got back...
This assumes that you can detect when a message is encrypted; if not (see my above post), you have to wade through millions of people who happen to be sending a lot more email than they are receiving.
Possibly desireable, but not necessary; techniques exist which allow embedded information to survive through compression, and even destruction of large chunks (half or more) of an image. Read up on "digital watermarking" -- I know Kodak has done a lot of research here.
Digital correspondence can always be buried in noise. A message could (for example) be written, heavily encrypted, and then embedded into a JPEG image in such a way that it would be undetectable without the right extraction software.
Does the government even have the processing power to do keyword searches on all the email that it monitors? Much less checking for the presence of heavily embedded messages...
I believe that any measures that the gov't takes to aid in electronic surveillence will only compromise our liberties.
I strongly agree with this. Anyone who has done research into artificial neural networks knows that the mathematics required to simulate them grows in complexity exponentially as the number of neurons in the network grow. We do not even posess the mathematical tools today to comprehensively model a network of ten neurons with a relatively simple activation function, unless we discretize the simulation in time.
I have done research which explored the presence of chaos in ANNs, and found that even such simple networks exhibit massive variances in outputs when very small changes are made to the parameters of their activation functions. And neuroscience is just scratching the surface of how a real neuron is activated electrochemically.
I think that the complexity of our own brains will push Kurzweil's predictions back at least a few centuries. But I'll save a copy of the interview for nostalgia's sake in 2050.