Don't forget the US shipments of equipment and supplies (especially dried eggs, nicknamed "Roosvelt's eggs") towards the Eastern Front.
All US shipments amounted to 4-7% of Soviet own production. Besides, USSR was paying for them with gold.
And I have no doubt that the US help, military and material, saved many thousands of lives. I just wanted to emphasize that saying 'US saved Europe from Nazis' is a gross overstatement.
it's clear that those who were there -- living or dead, friend or foe -- deserve too a lot of respect.
If France, sorry, the EU wants its copy of GPS, the U.S. will be ine with it. Until it's used to attack the U.S. At that point, it will cease to exist.
Does it means you are about to get rid of Boeing the company? Their products were used in terrorist attacks on US, after all.
those Americans are the menace, and yet those Americans saved Democracy in Europe twice in the 20th Century !
No, only once. US participation in defeating of Nazis, while significant, was miniscule compared to the scale of events on Eastern Front. Ironically, Europe was saved from Hitler by totalitarian Stalinist regime.
The technology field now consists primarily of certified-morons and foreign coders. What happened to the guys who started modern computing on its way back during World War II when they cracked the German enigma encryption by creating the computer? Where have all the talented white men gone?
Nono, Proud American, you should ask yourself where have all the talented men came from?
For your interest, Enigma code was cracked by a bunch of foreigners (some Brits and Poles).
Did you consider to add introspection mechanisms to your language, or is it intended to reason on 'concrete', outside-world level only? From your paper it seems you took the approach of simplest design (e.g. by using intermediate agents in communication vs. more elaborate interaction language). In my limited experience (I have my pet discovery system, also very simplistic), minimalist architecture encourages reflection and meta-reasoning, while complicating human inspection of system state.
Another thing is temporary K-lines. Do they essentially represent context information of a situation? And the whole K-lines + state mirrors combination resembles prototype-based object systems, where instances are clones of respective classes.
(Disclaimer: I just skimmed through your article, so maybe my understandings are wrong.)
No, I don't have a name for it. Any suggestions?:-)
Well, choosing a language name is prerogative of the author:)
Lisp is multi-paradigm language. Its OO system far exceeds the one of Java in elegance, flexibility and expressive power.
How many verison sof Lisp are there? Ever try to write a Lisp program that could runb in all implementations of Lisp.
ANSI Common Lisp is the dominant Lisp today, with numerous free and commerical implementations. Writing portable applications in it is certainly easier than, say, in C.
Every try to read a Lisp program with twenty nested function calls?
You don't need to nest 20 function calls very often (I believe I never did that). Lisp-aware source editors conveniently indent the source and match parens for you, so this is not an issue. The argument is on par with "ever tried to read a Java program written in single line"?
Whether we percieve the system as being inconsistent is irrelevant.
Percieve? I don't care about perceptions, we speak of formal systems here.
One can very well program a broken version of Peano arithmetic that would be inconsistent or incomplete. (The usefulness of that particular arithmetic not considered).
But as far as I know nobody prior to me has really tried to make K-lines, polynemes, pronomes, frames, etc., and hook them all together, as described in "Society of Mind".
How far you progressed with implementing the actual k-line language (btw, have it got a name)?
The new systems are doing some interesting things relative to the old symbolic AI systems (though they do have the advantage of 20 years of Moore's law to help them).
I maintain that they don't even do any truly novel things. Recognition and gradient search algorithms, while less refined, were already known 20 years back.
But the most irritating trend for me is that some people go into ANNs in hope that after tinkering with thresholds and topologies they'll come up with a magic box capable to write novels. They usually get stuck a problem like implementing XOR, but in process manage to piss over symbolic AI research achivements once again.
You have smart toys ala Aibo, and smart systems ala Eliza, and a lot of people is working towards creating smarter toys and smarter systems, but the real breaktrough will come when somebody manages to create the dumbest system possible.
Couldn't agree more. If someone will manage to create a GOFAI system, I am sure its core will fit in a few pages worth of source code.
How do you propose we build an inconsistent machine? Computers are absolutely dependent on the consistent processing of logical rules.
You can very well create an inconsistent system within consistent framework. Consider e.g. genetic algorhithms that mutate, crossover and substitute random pieces of code, then evaluate them using fitness function. You get a lot of crappy code this way, but what matters is the general direction of system development.
There are many authors that have written and demonstrated that the brain probably doesn't function as a mass of context-free predicate logic rules -- including my favorite, Hubert Dreyfus.
Dreyfus argument is old, and its rebuttals are well-known. Consider that symbolic systems are not limited to context-free predicate logic.
The progress of AI is uncertain, but it is certain that there's no future for symbolic logic AI.
It is not certain for me.
Both connectionist and symbolic approaches may succeed if given enough time. However, I think that obsession with neural nets of many people here is of the same nature that obsession of numerous early aviation enthusiasts with wind-flipping devices. Certainly you can mimic mechanics of nature with some effort, but there are usually better ways to do the job.
Re:Minsky only has himself to blame.
on
AI Going Nowhere?
·
· Score: 2, Informative
Promising subfields like perceptrons were intentionally quashed by him...
His take at perceptrons was vaild and well-founded. They indeed suffered from linear separability problem.
Roger Penrose argues that this is due to the fundemental Turing-style restrictions that we place on our notion of computing.
IIRC, Penrose argument boils down to the statement that no consistent and complete formal system can possess intelligence. Even if that one is true (and it is still merely an opinion), nothing prevents us from building inconsistent or incoplete systems. It may sound weird, but e.g. human thinking process is known to be remarkably inconsitent.
Search google for "liquid state machine". These researches are making progress novadays, not Minsky.
You seem to overestimate progress of artificial neural nets development. They show some impressive results in limited domains (mainly recognition and good old gradient search problems), but still have too many drawbacks to be called 'superior' to symbolic processing systems. And yes, I do know of liquid state machine, as of Boltzman machine, backpropagation nets, Hopfield networks etc.
Too many developers think it's beneath them to write good doco, example progs, tutorials, clear easily-learned APIs and clear meaningful comments in the code.
Then again, many FS/OSS developers do comment and document their code. Take a look, say, at Emacs.
Besides, your observation applies equally to both Open Source and proprietary projects. Programmer's skill is orthogonal to his area of activity. But of course, you are less likely to spot bad code in proprietary software, for the lack of sources.
This was a recent development, which we both like, contrary to your earlier statement, that you disklike all of them.
Maybe I should clarify if it wasn't clear from the context: I don't like that a single, powerful state ditches the judgement of United Nations and gets to promoting its interests using military force.
I must point out, that you deliberately dropped my next sentence -- that such unwillingness to rely on skill (not sincerety!) and wisdom of polititians condemns millions to decades of sufferings under brutal dictatorships.
Yes, pretty much as my denial of vigilant justice condemns thousands to suffering from criminals.
I maintain, that in some circumstances, the risk of "slipping" is worth it...
I'd state that it depends on your point of view. When you look at it as a citizen of the wanna-be empire, maybe. But definitely not when you are dweller of a country with not-so-good relations to the superpower.
I know the gun is practically a sniper weapon, but it was meant to allow the infantry soldier to "reach further" than one armed with an AK.
This is just an organisational issue. In 60s it was decided that each Soviet army platoon should have an attached sniper, who's primary targets are frontline command ranks of an enemy. In a way, of course, it increases effective range of small caliber fire of infantry.
But besides that there are separate sniper units in Russian military. SVD is still their primary weapon.
I can tell you some about the differences in build quality that make a sniper weapon.
I never claimed that SVD is the best sniper tool around. However, you can get only so far when you have to design cheap enough rifle for field use by drafted soldiers in multi-millon army. Finer devices would degrade pretty quick in hands of an IQ 80 owner crawling in mud.
I'd like to point out that although the SVD has "sniper" in its name, it is not really a sniper weapon.
It is really a sniper weapon. It always comes equipped with long-range optical sniper sight, and is normally loaded with more precise sniper ammo (that one has different kind of powder and bullet). Its holder's military designation in Soviet/Russian army is 'sniper', and they go through a different line of training than other infantrymen.
In Polish terms, a rifle is whatever shoots rifle ammo, and carbine is whatever shoots carbine ammo (less powerful).
Carabine is a well-known term for lighter rifles with shorter barrels, and it goes back to early 1800s, when it became a common cavalry weapon. Maybe it is different in Poland though.
I would be reaaaaally surprised if this fit a standard AK-47, as it is an SVD (Russian infantry rifle, as opposed to the AK, which is in fact a carbine, although called an assault rifle) mag.
SVD is an acronym from "Snaiperskaya Vintovka Dragunova" (Dragunov sniper rifle). It is clearly not a carabine.
Also, the term 'assault rifle' is absent from the official Russian military lexicon. I have no idea who would call SVD an assault rifle.
You are correct though that it was not an AK mag depicted.
Slashdot Mirror
Don't forget the US shipments of equipment and supplies (especially dried eggs, nicknamed "Roosvelt's eggs") towards the Eastern Front.
All US shipments amounted to 4-7% of Soviet own production. Besides, USSR was paying for them with gold.
And I have no doubt that the US help, military and material, saved many thousands of lives. I just wanted to emphasize that saying 'US saved Europe from Nazis' is a gross overstatement.
it's clear that those who were there -- living or dead, friend or foe -- deserve too a lot of respect.
Indeed.
If France, sorry, the EU wants its copy of GPS, the U.S. will be ine with it. Until it's used to attack the U.S. At that point, it will cease to exist.
Does it means you are about to get rid of Boeing the company? Their products were used in terrorist attacks on US, after all.
those Americans are the menace, and yet those Americans saved Democracy in Europe twice in the 20th Century !
No, only once. US participation in defeating of Nazis, while significant, was miniscule compared to the scale of events on Eastern Front. Ironically, Europe was saved from Hitler by totalitarian Stalinist regime.
The technology field now consists primarily of certified-morons and foreign coders. What happened to the guys who started modern computing on its way back during World War II when they cracked the German enigma encryption by creating the computer? Where have all the talented white men gone?
Nono, Proud American, you should ask yourself where have all the talented men came from?
For your interest, Enigma code was cracked by a bunch of foreigners (some Brits and Poles).
While they are at it, could they also made worm install some simple firewall and anti-viral software at user's marchines?
..who misread the topic as 'Peacemaker'-like GPS device?
Come to think of it, a Colt-shaped GPS reciever would look impressive..
..to the term 'core dump'.
A couple of questions:
:-)
:)
Did you consider to add introspection mechanisms to your language, or is it intended to reason on 'concrete', outside-world level only? From your paper it seems you took the approach of simplest design (e.g. by using intermediate agents in communication vs. more elaborate interaction language). In my limited experience (I have my pet discovery system, also very simplistic), minimalist architecture encourages reflection and meta-reasoning, while complicating human inspection of system state.
Another thing is temporary K-lines. Do they essentially represent context information of a situation? And the whole K-lines + state mirrors combination resembles prototype-based object systems, where instances are clones of respective classes.
(Disclaimer: I just skimmed through your article, so maybe my understandings are wrong.)
No, I don't have a name for it. Any suggestions?
Well, choosing a language name is prerogative of the author
I have programmed in Lisp.
Probably not a lot.
Lisp is a functional langauge.
Lisp is multi-paradigm language. Its OO system far exceeds the one of Java in elegance, flexibility and expressive power.
How many verison sof Lisp are there? Ever try to write a Lisp program that could runb in all implementations of Lisp.
ANSI Common Lisp is the dominant Lisp today, with numerous free and commerical implementations. Writing portable applications in it is certainly easier than, say, in C.
Every try to read a Lisp program with twenty nested function calls?
You don't need to nest 20 function calls very often (I believe I never did that). Lisp-aware source editors conveniently indent the source and match parens for you, so this is not an issue. The argument is on par with "ever tried to read a Java program written in single line"?
Whether we percieve the system as being inconsistent is irrelevant.
Percieve? I don't care about perceptions, we speak of formal systems here.
One can very well program a broken version of Peano arithmetic that would be inconsistent or incomplete. (The usefulness of that particular arithmetic not considered).
But as far as I know nobody prior to me has really tried to make K-lines, polynemes, pronomes, frames, etc., and hook them all together, as described in "Society of Mind".
How far you progressed with implementing the actual k-line language (btw, have it got a name)?
If you have some specific references to other rebuttals please post them, they'd be most welcome on my reading list.
Maybe not quite a rebuttal, but McCarthy home page contains descriptions of nonmonotonic reasoning systems which, while logical, are not context-free.
The new systems are doing some interesting things relative to the old symbolic AI systems (though they do have the advantage of 20 years of Moore's law to help them).
I maintain that they don't even do any truly novel things. Recognition and gradient search algorithms, while less refined, were already known 20 years back.
But the most irritating trend for me is that some people go into ANNs in hope that after tinkering with thresholds and topologies they'll come up with a magic box capable to write novels. They usually get stuck a problem like implementing XOR, but in process manage to piss over symbolic AI research achivements once again.
You have smart toys ala Aibo, and smart systems ala Eliza, and a lot of people is working towards creating smarter toys and smarter systems, but the real breaktrough will come when somebody manages to create the dumbest system possible.
Couldn't agree more. If someone will manage to create a GOFAI system, I am sure its core will fit in a few pages worth of source code.
How do you propose we build an inconsistent machine? Computers are absolutely dependent on the consistent processing of logical rules.
You can very well create an inconsistent system within consistent framework. Consider e.g. genetic algorhithms that mutate, crossover and substitute random pieces of code, then evaluate them using fitness function. You get a lot of crappy code this way, but what matters is the general direction of system development.
There are many authors that have written and demonstrated that the brain probably doesn't function as a mass of context-free predicate logic rules -- including my favorite, Hubert Dreyfus.
Dreyfus argument is old, and its rebuttals are well-known. Consider that symbolic systems are not limited to context-free predicate logic.
The progress of AI is uncertain, but it is certain that there's no future for symbolic logic AI.
It is not certain for me.
Both connectionist and symbolic approaches may succeed if given enough time. However, I think that obsession with neural nets of many people here is of the same nature that obsession of numerous early aviation enthusiasts with wind-flipping devices. Certainly you can mimic mechanics of nature with some effort, but there are usually better ways to do the job.
Promising subfields like perceptrons were intentionally quashed by him...
His take at perceptrons was vaild and well-founded. They indeed suffered from linear separability problem.
Roger Penrose argues that this is due to the fundemental Turing-style restrictions that we place on our notion of computing.
IIRC, Penrose argument boils down to the statement that no consistent and complete formal system can possess intelligence. Even if that one is true (and it is still merely an opinion), nothing prevents us from building inconsistent or incoplete systems. It may sound weird, but e.g. human thinking process is known to be remarkably inconsitent.
Search google for "liquid state machine". These researches are making progress novadays, not Minsky.
You seem to overestimate progress of artificial neural nets development. They show some impressive results in limited domains (mainly recognition and good old gradient search problems), but still have too many drawbacks to be called 'superior' to symbolic processing systems. And yes, I do know of liquid state machine, as of Boltzman machine, backpropagation nets, Hopfield networks etc.
His keyboard '5' key is probably broken. This conjecture is supported by lack of the digit '5' in all the percentages he listed.
Too many developers think it's beneath them to write good doco, example progs, tutorials, clear easily-learned APIs and clear meaningful comments in the code.
Then again, many FS/OSS developers do comment and document their code. Take a look, say, at Emacs.
Besides, your observation applies equally to both Open Source and proprietary projects. Programmer's skill is orthogonal to his area of activity. But of course, you are less likely to spot bad code in proprietary software, for the lack of sources.
This was a recent development, which we both like, contrary to your earlier statement, that you disklike all of them.
Maybe I should clarify if it wasn't clear from the context: I don't like that a single, powerful state ditches the judgement of United Nations and gets to promoting its interests using military force.
I must point out, that you deliberately dropped my next sentence -- that such unwillingness to rely on skill (not sincerety!) and wisdom of polititians condemns millions to decades of sufferings under brutal dictatorships.
Yes, pretty much as my denial of vigilant justice condemns thousands to suffering from criminals.
I maintain, that in some circumstances, the risk of "slipping" is worth it...
I'd state that it depends on your point of view. When you look at it as a citizen of the wanna-be empire, maybe. But definitely not when you are dweller of a country with not-so-good relations to the superpower.
I know the gun is practically a sniper weapon, but it was meant to allow the infantry soldier to "reach further" than one armed with an AK.
This is just an organisational issue. In 60s it was decided that each Soviet army platoon should have an attached sniper, who's primary targets are frontline command ranks of an enemy. In a way, of course, it increases effective range of small caliber fire of infantry.
But besides that there are separate sniper units in Russian military. SVD is still their primary weapon.
I can tell you some about the differences in build quality that make a sniper weapon.
I never claimed that SVD is the best sniper tool around. However, you can get only so far when you have to design cheap enough rifle for field use by drafted soldiers in multi-millon army. Finer devices would degrade pretty quick in hands of an IQ 80 owner crawling in mud.
I'd like to point out that although the SVD has "sniper" in its name, it is not really a sniper weapon.
It is really a sniper weapon. It always comes equipped with long-range optical sniper sight, and is normally loaded with more precise sniper ammo (that one has different kind of powder and bullet). Its holder's military designation in Soviet/Russian army is 'sniper', and they go through a different line of training than other infantrymen.
In Polish terms, a rifle is whatever shoots rifle ammo, and carbine is whatever shoots carbine ammo (less powerful).
Carabine is a well-known term for lighter rifles with shorter barrels, and it goes back to early 1800s, when it became a common cavalry weapon. Maybe it is different in Poland though.
I would be reaaaaally surprised if this fit a standard AK-47, as it is an SVD (Russian infantry rifle, as opposed to the AK, which is in fact a carbine, although called an assault rifle) mag.
SVD is an acronym from "Snaiperskaya Vintovka Dragunova" (Dragunov sniper rifle). It is clearly not a carabine.
Also, the term 'assault rifle' is absent from the official Russian military lexicon. I have no idea who would call SVD an assault rifle.
You are correct though that it was not an AK mag depicted.