No, actually they don't. Fundamentally it all comes back to the performer and the performance. Film fundamentally changed a performers placed in the world. Prior to film anyone who wanted to see their act had to see them. After film their act could be recorded once and played forever, especially later when sound and color were added. TV changed things again in that people no longer had to go to a movie theatre. And now the digital media revolution has once again changed the dynamic.
In each case in the past the market eventually came up with a system that left all sides feeling reasonably fairly treated. Legislation will in no way help the process that will need to take place this time to make that happen again.
I write songs in my spare time. I haven't sold any yet, but say I want to. Many of the laws that have been put in place will make it difficult for me to do so without somehow getting buyin from one of the current big players in this arena. And don't think for an instant that this is just an unintended side effect. The whole point of these laws is to keep the media giants in business, largely unchanged and thriving in the face of a radical technological shift that should at the very least cause them to change significantly...
Imagine if the vaudeville and stage actors had gotten together back in the early 20th century and gotten motion pictures outlawed. Or if actresses and actors who had horrible speaking voices had gotten talkies outlawed in the late 1920's, etc.
The recent legal trend to try to hold back technological progress is disturbing to say the least. Its also stupid and futile in the long run.
This is basically the exact same pattern Microsoft followed in the early 90's in regards to the internet but in a slightly different guise. The goal is to box the people who use their OS in to using Microsoft and only Microsoft. We all recall how well it worked then and it is likely to work about as well this time.
By 2007 Microsoft will be porting all their applications to run under Linux and will be back to trying their "embrace, extend, attempt to strangle" stratagy.
Too bad our government is so spineless. Microsoft keeps doing the same things over and over again and will continue to do so as long the penalty continues to be as miniscule as it has been up until now.
How?
This would require the capacity for self deception. How would you program that and still end up with anything approaching a functional AI? And more importantly WHY would you do so?
My supposition is that it would require far more effort to create and AI that had the capacity for insanity then it would to create a perfectly sane one. Why am I wrong? (If I am...)
How? I maintain that it would be far more difficult (And undesirable to boot) to create a psychotic or neurotic AI. It would require a significant amount of additional code. Why would anyone bother (Even assuming it is possible?)
If you're an AI then you have to reduce an ambiguous statement to something concrete if you want to understand it and be able to take some sort of action. Making an AI that is identical to a human would I suspect be very very difficult. It would also be largely pointless since we have plenty of human beings and no need to build artificial ones. Having access to intelligences that were logical and incapable of self deception on the other hand would potentially be very useful in many applications.
I suppose it might be relavent to define what I mean by insanity. In retrospect it may have been a bad choice of words. "Irrational" would probably be a better choice of words.
If a person is presented with a rational argument that includes well documented and supported facts then there is still a good chance that they will reject it if it is counter to their currently held opinion on a particular topic. This to me is a mild form of insanity though the medical profession would as you noted disagree.
Thus my question/assertions was really that creating an irrational AI would be even more difficult then creating a purely rational one. And note when I use the word irrational here I am refering to the level of irrationality in your average ordinary human being.
My question has to do with sanity. Specifically is it possible for an AI to be insane? To elaborate, any artificial intelligence is going to require very sophisticated algorythms. These algorythms are going to likely have significant components focused on logical consistency as it is much easier to handle logically consistent concepts then the fuzzy ugly ones we humans deal with.
There is a language called lojban I believe that is completely unambiguous. If you were to translate human input into lojban as an intermediate step in having the AI handle input then you would end up with no ambiguity. The reduction in ambiguity would make it very difficult for the AI to misunderstand or deceive itself (Assuming the translation were correct). Since instany seems to be based in large part on the ability to self deceive the removal of self deception from the input along with the need to keep things as logical and self consistent as possible internally would tend to argue to me that insanity in a functional AI would be very unlikely.
The answer is simple. Change the economics of email. Do it in a gradual way and make sure it remains backwards compatible during the process. It would likely take several years to completely replace the existing infrastructure but it could be done.
Slashdot Mirror
No, actually they don't. Fundamentally it all comes back to the performer and the performance. Film fundamentally changed a performers placed in the world. Prior to film anyone who wanted to see their act had to see them. After film their act could be recorded once and played forever, especially later when sound and color were added. TV changed things again in that people no longer had to go to a movie theatre. And now the digital media revolution has once again changed the dynamic.
In each case in the past the market eventually came up with a system that left all sides feeling reasonably fairly treated. Legislation will in no way help the process that will need to take place this time to make that happen again.
I write songs in my spare time. I haven't sold any yet, but say I want to. Many of the laws that have been put in place will make it difficult for me to do so without somehow getting buyin from one of the current big players in this arena. And don't think for an instant that this is just an unintended side effect. The whole point of these laws is to keep the media giants in business, largely unchanged and thriving in the face of a radical technological shift that should at the very least cause them to change significantly...
Imagine if the vaudeville and stage actors had gotten together back in the early 20th century and gotten motion pictures outlawed. Or if actresses and actors who had horrible speaking voices had gotten talkies outlawed in the late 1920's, etc. The recent legal trend to try to hold back technological progress is disturbing to say the least. Its also stupid and futile in the long run.
This is basically the exact same pattern Microsoft followed in the early 90's in regards to the internet but in a slightly different guise. The goal is to box the people who use their OS in to using Microsoft and only Microsoft. We all recall how well it worked then and it is likely to work about as well this time.
By 2007 Microsoft will be porting all their applications to run under Linux and will be back to trying their "embrace, extend, attempt to strangle" stratagy.
Too bad our government is so spineless. Microsoft keeps doing the same things over and over again and will continue to do so as long the penalty continues to be as miniscule as it has been up until now.
How? This would require the capacity for self deception. How would you program that and still end up with anything approaching a functional AI? And more importantly WHY would you do so? My supposition is that it would require far more effort to create and AI that had the capacity for insanity then it would to create a perfectly sane one. Why am I wrong? (If I am...)
How? I maintain that it would be far more difficult (And undesirable to boot) to create a psychotic or neurotic AI. It would require a significant amount of additional code. Why would anyone bother (Even assuming it is possible?)
If a person is presented with a rational argument that includes well documented and supported facts then there is still a good chance that they will reject it if it is counter to their currently held opinion on a particular topic. This to me is a mild form of insanity though the medical profession would as you noted disagree.
Thus my question/assertions was really that creating an irrational AI would be even more difficult then creating a purely rational one. And note when I use the word irrational here I am refering to the level of irrationality in your average ordinary human being.
My question has to do with sanity. Specifically is it possible for an AI to be insane? To elaborate, any artificial intelligence is going to require very sophisticated algorythms. These algorythms are going to likely have significant components focused on logical consistency as it is much easier to handle logically consistent concepts then the fuzzy ugly ones we humans deal with. There is a language called lojban I believe that is completely unambiguous. If you were to translate human input into lojban as an intermediate step in having the AI handle input then you would end up with no ambiguity. The reduction in ambiguity would make it very difficult for the AI to misunderstand or deceive itself (Assuming the translation were correct). Since instany seems to be based in large part on the ability to self deceive the removal of self deception from the input along with the need to keep things as logical and self consistent as possible internally would tend to argue to me that insanity in a functional AI would be very unlikely.
The answer is simple. Change the economics of email. Do it in a gradual way and make sure it remains backwards compatible during the process. It would likely take several years to completely replace the existing infrastructure but it could be done.