Re:Definition of intelligence - it's most basic fo
on
Arguing A.I.
·
· Score: 1
Quite. I'd define intelligence in an a similar way, we recognise intelligence as the ability to manipulate models and get predictions that are close to reality. Someone who's world view differs significantly from observed reality we would consider insane, their results do not hold up. But its all about context. In the limited context of a chess game, then yes, Deep Blue is intelligent. But what we're talking about is 'human level intelligence' which is the ability to generate 'intelligent' results in a wide and undefined set of systems (after all, the real world isn't always predictable).
Re:Penrose and A.I.
on
Arguing A.I.
·
· Score: 2, Insightful
I cannot really get very excited by Penrose's foray into AI, he brings things to the table almost just because he can. Of course, he may turn out to be right to have done so, but, I can't help thinking he's just trying to justify some dualistic belief system he has.
Still, they are worth a read.
And the question of what intelligence IS, seems to be really the nub of the matter. How will we know we've created an AI if we don't even know what intelligence is?
It's a systems thing again, out of context, intelligence is not recognisable. Only in the context of a wider system (e.g. the world for human purposes) does intelligence become something we can even discuss.
Intelligence in humans arose to enable us to build better models and hence make better predictions, and is intimately related to our tribe/troop/pack social structure and the increasing complexity that it developed.
Intelligence then enabled even more complex structures and relationships, requiring greater intelligence to 'compete' in, and so on. I suspect that eventually the 'cost' of that intelligence became greater than the returns and it all levelled out to where we are today.
Re:Iain M. Banks - The Culture Series
on
Arguing A.I.
·
· Score: 1
If you want something a bit closer to home as it were, try David Gerrold's 'When H.A.R.L.I.E. Was One - Release 2.0', probably the best piece of fiction on AI.
Re:complexity of supercomputers approaching brain
on
Arguing A.I.
·
· Score: 1
You rather convieniently forgot about those 'thousands of connections' - these are vital to how the brain operates, so you'd need to figure out the i/o requirements - I reckon we're a LOT further away on that front...
Re:Von Neumann Architecture Can't Do It.
on
Arguing A.I.
·
· Score: 1
The assumption is that a digital computer can model any physical system given the right software. If that assumption holds, then you can model the brain on a digital computer. If it doesn't hold, all bets are off. My suspicion is that there are physical systems that cannot be modelled at any level beyond abstractions, but that the brain is not in that domain and hence is theoretically modelable. But the gap between theory and practice in this case is so vast...
Re:Chinese Rooms and Software Guys
on
Arguing A.I.
·
· Score: 1
And all that comes back to Turing's test, if you can't tell the difference, then is there a difference?
And as for "the mind is software that runs on the brain" - nope, no software, just brain. In a sense, there is no software in a computer (or rather, that's just a name refering to what is actually a pattern of electric charge). Dualism posits that 'mind' is some non-physical thing that must exist effectively outside of the universe. Strong AI researchers would posit something quite different, that 'mind' is just a state of a highly complex and entirely physical system.
In a way, a 'mind' is just a finite state machine, but, that's quite some finite! For all practical purposes we should really just think of it as an infinite state machine.
AI is possible (unless you go down the dualist road and invoke souls or simillar undefined and probably undefinable entities), our existance is proof enough. AI on a conventional digital computer is also almost certainly possible, but I highly doubt its practicable - how big a computer would you need to model a human brain?
The earliest example of Cyberspace I can think of off the top of my head is the Dr. Who story line 'The Deadly Assassin' broadcast in 1976, but I'm sure someone can come up with an earlier example.
If there was one thing Commodore-Amiga ever got right, it was producing a standard script driven installer with logging and dry-run features, selection of levels of competence, etc. Its one thing I really miss on other systems. The script langage was a bit nasty being loosley lisp based and full of brackets, but otherwise...
Unix/X is like DOS/Win9x, the windowing stuff is bolted ontop of the CLI environment. In the Mac, the windowing stuff is default, but there is no CLI. Anyone who remembers the Amiga wil know it used windows by default, but the initial interface was a CLI - IN A WINDOW. The standard startup initialised the Workbench graphical shell and closed the initial CLI but did not have to do this. I still miss this flexibility and having the CLI as an integrated part of the windowing environment. UNIX desktops come close though, but the windowing is still an afterthought.
I'm not sure how Be handles this, maybe someone can provide that info.
If there is a way for a program to figure out if it has access to a GUI, then it can present a suitable interface (although a CLI switch should be available to supress that). I guess that would provide a best-of-both-worlds interface, CLI users get words, GUI users get windows (unless they say otherwise).
Just my 2p worth...
I had most of this over 10 years ago with my Amiga, so one wonders what we've all been doing since then. Ok, so no memory management (not that important, well written and tested software won't crash often, where as badly written, bloated softare will, even with MM eg Windows), and unsophisticated IPC (shared memory), but everything else was there - microkernel OOP executive, modular librarie/devices, handlers (translators in the HURD speak). I wrote a mail server in ARexx using the IP handler exposed by AmiTCP in around 12 lines!
Oh well, no use living in the past (which appears to have actually been the future), so I remain interested in the HURD although I'd rather not have all that UNIX junk piled ontop. Hmm, wonder if someone could do AmigaOS ontop of this:)
Re:Family values are inverse of technological valu
on
Trigger Happy
·
· Score: 1
Yeah, but your parents have watched too much TV and had their brains rotted:-)
Re: Amiga
Thats knida what I thought, and even that wasn't new. The only stuff I'm not sure about is the virtual/physical address mapping stuff, something the Amiga never had to worry about not having virtual memory (ignoring 3rd party apps).
Its not that clear what these patents are about, several of them seem to repeat themselves a lot. Probably just shows what a mess the USPO is in!
There's an interesting article in this week's New Scientist about actually making use of multipath
reflections to enable more bandwidth to be squeezed out of a given transmission band by simply increasing the number of transmitters and receivers (well, OK, not simply, but you get the gist). Combined with an advanced modulation technique like this...
That inflexibility sounds a lot like the British in WW1 where the class ensured that the officers were simply the higher classes, rather than the most competent. The working class soldiers also held their officers in some esteem due to them being further up the class system. The carnage that resulted pretty much destroyed the class system in Britain, or at least reduced it to a mere shadow of its former self. And the working classes would never again trust the ruling elite to the extent they did before the war.
So, whilst culture shapes language (and vive-verca), I suspect culture was the main factor here, not any ambiguity in the language.
Unix/Linux is based on 30 year old code and concepts. Some of those concepts were pretty novel at the time, many are still good ideas in theory. DLL hell on Windows is a problem with software installations, Microsoft themselves are guiltier than most with newer DLLs that are not backward compatible (hence the strict interface rules for COM objects). Perhaps th Open Source movement should work on developing a modern OS round a micro-kernel with other services exposed as object based components. Even the 15 year old Amiga had a more advanced OS in many respects than what we all currently use (even if it lacked in many areas). The new Amiga stuff sounds interesting mind you. Just my tupence worth:)
Slashdot Mirror
Quite. I'd define intelligence in an a similar way, we recognise intelligence as the ability to manipulate models and get predictions that are close to reality. Someone who's world view differs significantly from observed reality we would consider insane, their results do not hold up. But its all about context. In the limited context of a chess game, then yes, Deep Blue is intelligent. But what we're talking about is 'human level intelligence' which is the ability to generate 'intelligent' results in a wide and undefined set of systems (after all, the real world isn't always predictable).
I cannot really get very excited by Penrose's foray into AI, he brings things to the table almost just because he can. Of course, he may turn out to be right to have done so, but, I can't help thinking he's just trying to justify some dualistic belief system he has.
Still, they are worth a read.
And the question of what intelligence IS, seems to be really the nub of the matter. How will we know we've created an AI if we don't even know what intelligence is?
It's a systems thing again, out of context, intelligence is not recognisable. Only in the context of a wider system (e.g. the world for human purposes) does intelligence become something we can even discuss.
Intelligence in humans arose to enable us to build better models and hence make better predictions, and is intimately related to our tribe/troop/pack social structure and the increasing complexity that it developed.
Intelligence then enabled even more complex structures and relationships, requiring greater intelligence to 'compete' in, and so on. I suspect that eventually the 'cost' of that intelligence became greater than the returns and it all levelled out to where we are today.
If you want something a bit closer to home as it were, try David Gerrold's 'When H.A.R.L.I.E. Was One - Release 2.0', probably the best piece of fiction on AI.
You rather convieniently forgot about those 'thousands of connections' - these are vital to how the brain operates, so you'd need to figure out the i/o requirements - I reckon we're a LOT further away on that front...
The assumption is that a digital computer can model any physical system given the right software. If that assumption holds, then you can model the brain on a digital computer. If it doesn't hold, all bets are off. My suspicion is that there are physical systems that cannot be modelled at any level beyond abstractions, but that the brain is not in that domain and hence is theoretically modelable. But the gap between theory and practice in this case is so vast...
And all that comes back to Turing's test, if you can't tell the difference, then is there a difference?
And as for "the mind is software that runs on the brain" - nope, no software, just brain. In a sense, there is no software in a computer (or rather, that's just a name refering to what is actually a pattern of electric charge). Dualism posits that 'mind' is some non-physical thing that must exist effectively outside of the universe. Strong AI researchers would posit something quite different, that 'mind' is just a state of a highly complex and entirely physical system.
In a way, a 'mind' is just a finite state machine, but, that's quite some finite! For all practical purposes we should really just think of it as an infinite state machine.
AI is possible (unless you go down the dualist road and invoke souls or simillar undefined and probably undefinable entities), our existance is proof enough. AI on a conventional digital computer is also almost certainly possible, but I highly doubt its practicable - how big a computer would you need to model a human brain?
The earliest example of Cyberspace I can think of off the top of my head is the Dr. Who story line 'The Deadly Assassin' broadcast in 1976, but I'm sure someone can come up with an earlier example.
If there was one thing Commodore-Amiga ever got right, it was producing a standard script driven installer with logging and dry-run features, selection of levels of competence, etc. Its one thing I really miss on other systems. The script langage was a bit nasty being loosley lisp based and full of brackets, but otherwise...
Unix/X is like DOS/Win9x, the windowing stuff is bolted ontop of the CLI environment. In the Mac, the windowing stuff is default, but there is no CLI. Anyone who remembers the Amiga wil know it used windows by default, but the initial interface was a CLI - IN A WINDOW. The standard startup initialised the Workbench graphical shell and closed the initial CLI but did not have to do this. I still miss this flexibility and having the CLI as an integrated part of the windowing environment. UNIX desktops come close though, but the windowing is still an afterthought. I'm not sure how Be handles this, maybe someone can provide that info. If there is a way for a program to figure out if it has access to a GUI, then it can present a suitable interface (although a CLI switch should be available to supress that). I guess that would provide a best-of-both-worlds interface, CLI users get words, GUI users get windows (unless they say otherwise). Just my 2p worth...
I had most of this over 10 years ago with my Amiga, so one wonders what we've all been doing since then. Ok, so no memory management (not that important, well written and tested software won't crash often, where as badly written, bloated softare will, even with MM eg Windows), and unsophisticated IPC (shared memory), but everything else was there - microkernel OOP executive, modular librarie/devices, handlers (translators in the HURD speak). I wrote a mail server in ARexx using the IP handler exposed by AmiTCP in around 12 lines! Oh well, no use living in the past (which appears to have actually been the future), so I remain interested in the HURD although I'd rather not have all that UNIX junk piled ontop. Hmm, wonder if someone could do AmigaOS ontop of this :)
Yeah, but your parents have watched too much TV and had their brains rotted :-)
Re: Amiga Thats knida what I thought, and even that wasn't new. The only stuff I'm not sure about is the virtual/physical address mapping stuff, something the Amiga never had to worry about not having virtual memory (ignoring 3rd party apps). Its not that clear what these patents are about, several of them seem to repeat themselves a lot. Probably just shows what a mess the USPO is in!
There's an interesting article in this week's New Scientist about actually making use of multipath reflections to enable more bandwidth to be squeezed out of a given transmission band by simply increasing the number of transmitters and receivers (well, OK, not simply, but you get the gist). Combined with an advanced modulation technique like this...
That inflexibility sounds a lot like the British in WW1 where the class ensured that the officers were simply the higher classes, rather than the most competent. The working class soldiers also held their officers in some esteem due to them being further up the class system. The carnage that resulted pretty much destroyed the class system in Britain, or at least reduced it to a mere shadow of its former self. And the working classes would never again trust the ruling elite to the extent they did before the war. So, whilst culture shapes language (and vive-verca), I suspect culture was the main factor here, not any ambiguity in the language.
Unix/Linux is based on 30 year old code and concepts. Some of those concepts were pretty novel at the time, many are still good ideas in theory. DLL hell on Windows is a problem with software installations, Microsoft themselves are guiltier than most with newer DLLs that are not backward compatible (hence the strict interface rules for COM objects). Perhaps th Open Source movement should work on developing a modern OS round a micro-kernel with other services exposed as object based components. Even the 15 year old Amiga had a more advanced OS in many respects than what we all currently use (even if it lacked in many areas). The new Amiga stuff sounds interesting mind you. Just my tupence worth :)