Let me get this straight... he wants hardware that will detect all possible programs that will copy digital media...
So, from a theoretical computer science point of view, he wants a Turing machine that will recognize all Turing machines that compute a fixed function f. That sounds remarkably like a problem that is equivalent (by reduction) to the halting problem for Turing machines... Oh, did we mention that the halting problem is unsolvable??
But hey, if *Mr. Valenti* says so, it *must* be possible. After all, everyone knows that you can simply legislate away fundamental laws of mathematics...
Whats next? Valenti proposing that we set Pi equal to 3.0 to simplify calculations?
>Is it just me, or do others wish that they could get the fan count down in their boxes?
Its not just you =) I recently went through the horrifying experience of upgrading my home machine from a PIII-450 to an Athlon XP (eXtra Pyrotechnics?).
On the upside, the money I invested in a new case and all those fans is more than offset by the savings I get since I no longer have to heat my apartment. There is the small matter of the power bills though...
The US "owns" the *whole* internet, and paid for it? Wow.. am I ignorant. I guess it makes sense though, since Al Gore 'invented' it.
As far as I know, ARPAnet and the TCP/IP protocol stack were invented by scientists working in America (but you might want to go back and check all their nationalities... you might find a few suprises!). By extension the early internet was mainly in the US. But hey, have you ever heard of JANet (Joint Academic Network)? You know the early packet-switching network that the BRITISH had?
Yup, true story. 'Back in the day', there was a single link between what you so quaintly refer to as 'the whole internet' (e.g. the US network that grew out of ARPAnet) and JANet. You had to telnet to a host out on the US east cost (name escapes me at the moment, it was a long time ago) and then enter the JANet address you wanted.
I seem to remember that JANet address were 'backwards' too (e.g. org.slashdot.www )
As far as I know, JAnet was designed and built pretty much autonomously by the British. I don't ever recall hearing about American's desigining it, or paying for it. In fact, I'm pretty sure it was based on X.25 too.
Anyway, the point is that several other countries had their own packet switched networks that eventually got hooked together to create the 'internet'. As the lowest common denominator its true that TCP/IP become the de facto standard, and yes this is an American invention.
But the honest truth is that the U.S. "paid for and own" the internet about as much as Al Gore invented pants.
>crist ever wonder why we have hospitaols that are way better than any crap you have?
Okay, I'm prepared to hear out your opinion if you can provide me with some actual *fact* to back up your ranting. Trying to make such a sweeping arguement with a single anecdote doesn't lend credibility to your argument.
Can you perhaps tell us what year of Med school you're in? You make several statements of opinion, but to be honest, you kinda have to show that you're an expert in the field before personal opinion becomes a valid arguing point.
Lets take a look at research productivity, shall we? Certainly this isn't the best measure of how good a health care system is, but its one thats very easy to quantify. And really, if there is enough time and money to do research, one assumes the patients are at least not suing for malpractice.
Grab a fistful of medical journals at your local university library. Count how many papers have first authors at Canadian institutions. Count how many at American institutions. Now factor in that our population is 1/10 the size of yours.
I'm not going to tell you what to think, I'll let you draw your own conclusions.
The hard reality is that if you're rich, there is no question US healthcare is better. The US is a capitalist county and if you work hard and are successfull, you can have *whatever* you want. Including the worlds best healthcare.
On the other hand Canada, which is really rather socialist, is a MUCH better place to be if you're poor. In fact, even if you're a street person, you'll receive the same level of care as a CEO worth millions of dollars (in theory).
If social responsibility is your idea of a "cluster-F**k", then I guess you have a point. If, on the other hand, you think that maybe as human beings we can evolve beyond a selfish ME FIRST attitude; I think you have to re-examine the facts.
Note that I'm not out and out supporting socialist healthcare policies here... I'm not convinced its the right way to go. India is a *very* socialist country and to be honest, I would not want to live there.
Anyway the point was that straight-out "x is better than y" comparisons of healthcare are really bogus. US > CAN if you're rich; CAN > US if you're poor. If you're middle class, its not immediately apparent which is better.
The thing is, you'd never put an embargo on us because we supply you with such useful things as:
- Timber
- Grain
- Uranium
- Electricity (get your facts straight)
- Funny comedians
- oh yes, and a little thing called FRESH WATER
You really *do* need our natural resources, and for the record we have more oil here than in all of Texas. Its just that its in tar sands and really expensive to refine; but if we had no other choice, we could get along quite nicely.
It was ludicrous claims like this about "the magical wonders of AI" that ended up with AI winter back in the 80's...
>I don't know what this guy's smoking.
A big pipe of ignorance... anyone who's taken a basic intro to AI course knows the crushing disillusion one experiences when one realizes AI is really just a bag of tricks.
Sophisticated and interesting tricks to be sure, but at the end of the day isn't it much more magical to believe the computer is really *thinking* about that game of chess... it sure takes the fun out of it once you learn about alpha-beta pruning, etc.
I wouldn't be surprised if this guy also thinks that David Copperfield really *can* teleport objects and levitate. I guess Clarke's law applies here, and to this particular individual, AI is a 'sufficiently advanced technology'... thus rendering it indistinguishable from magic and suitable for wild-eyed predictions.
>under current understood physics... Same goes for the rest.
I respectfully disagree... Godel's incompleteness theorem holds up even under alternative laws of physics. Though the whole P/NP thing doesn't (there was a paper on the xxx.lanl preprint servers a few years back showing that if quantum mechanics is non-linear then P=NP as a consequence. Good reading.)
You make some excellent points, but I disagree slightly with the following statement:
>The fact that there are corporations out there that
>are willing to give money to these athletes is
>enough for me to want to support those corporations
>so I don't mind knowing who the olympic sponsers
>are of the national team.
For those corporations "Olypmic Sponsorship" == "Tax-deductable advertising". Think about it. In your argument you mentioned Roots, so you were thinking about Roots clothing, and then I thought about Roots, and any other Canadian who read it thought about Roots too.
Even better... most of us had warm and fuzzy (no pun intended;) ) thoughts about Roots too...
On the other hand, it is nice that they sponsor our atheletes... altruistic or not. You made the excellent point that, regardless of intent, the atheletes still get sponsored.
Now heres something we can all agree on:
>I am totally surrounded by 24 hr olympics here on
>a sat. afternoon between my television, radio, and
>internet connection.. and comfy for 2 weeks of the
>hype.. I love it
Damn straight! GO Canada!
Thanks for dragging me out of my anti-capitalist cynacism;)
a) competition between amateur atheletes
b) the spirit of global friendship
c) Incredible advertising and product placement opportunities
d) making as much money as you can fit down your bloated gullet
e) all of the above
I guess I put option e) in there because I'm not a complete cynic yet; but lets face it... the games *are* commerical...
Its a good point, and you're correct that Canadian law *should* in theory have allowed this. If you want an example of what happens in practice though, check out the "icravetv.com" story and the resulting effect...
http://www.newsbytes.com/news/01/172941.html
Once again, it doesn't matter who's right... it matters who has more money. As long as you can buy... I mean *lobby* the CRTC and/or a good number of MPs, you can just get the offending law changed.
I think the following statement in the press release pretty much says it all:
>We perceive this advancement as a significant
>breakthrough to the historical limitations of
>digital communications as it was originally
>detailed by
>Dr. Claude Shannon in his treatise on Information
>Theory."
How about algorithmic information theory? Kolmogorov, Solomonov, Chaitin? The statement above indicates that the most recent word on compression is an old Bell Labs tech report by Claude Shannon... not to put Shannon down, that work *is* a landmark, but there has certainly been more work done since.
Try compressing the number Pi using Shannons theory... you can't do it. On the other hand, using Kolmogorov complexity, you can compress it quite nicely.
The fact that this statement appears in the press release seems to indicate a great deal of ignorance on the part of this corporations researchers. Part of any good research program is to familiarize yourself with previous work done in the field... and AIT is *not* some obscure backwater idea... there are several conferences on this topic every year and just about every CS graduate student has seen at least Kolmogorov complexity.
This is a pretty serious credibility robber. (Not to mention that from a mathematical standpoint, compressing totally random data is impossible under our current axioms... so if we *can* compress completely random data... its time for a new theory of the foundations of mathematics. At the risk of sounding dogmatic: do you *really* think some dot-com startup is capable of this?
Perhaps they are, but I'm going to need to see the proofs written up nice and formally before I run out and buy snake-oi... I mean *stock*.)
Its an interesting proposition, and I thought I'd chime in with my viewpoint because I think I have a (semi) qualified opinion here.
I work at a university which has a significant number of graduate students who received their undergraduate education from institutions in mainland China. Moreover, we literally receive hundreds of applications from such students every year. I served on the admissions committee for the graduate program last year and had the opportunity to review transcripts and CVs for many of these individuals.
I am of the opinion that what constitutes a CS education in China is radically different from what constitutes a CS education in North America and Europe. I saw several applicants who received degrees in computer science, yet the only courses I could find on their transcripts that were even vaguely related were "Visual Basic Programming I, II and III" and "Microsoft Access".
No theory courses, no programming languages, no algorithms, no data structures, no design, no software engineering. There were, however, mandatory 'Physical Education' and 'Marxist Economic Theory' courses. (I really feel for these students... I'm not sure how thrilled I would've been to have mandatory "gym class" in University!)
Unfortunately this is not a just a few cases; this was the norm (at least for the few hundred applications I've personally seen).
My overall impression then, is that these applicants probably are qualified to hack out VB code. In fact, their education seems to be setting them for trade jobs as coders. Unfortunately, whats lacking is any sort of rigourous training in basic computer *science* and software engineering. In the long run, one of three things can come of this: these individuals will educate themselves and become productive coders; someone else (i.e. their employers) will have to pay to educate them on the missing material; or, sadly, they simply will never be very effective software engineers. Since the people who will choose path #1 are few and far between, I'm not sure India has anything to worry about yet.
I think China needs to take a serious look at how they are educating their 'Computer Scientists' if they are serious about becoming a world software power. Its certainly within their grasp, I'm just not convinced they have the system in place to achieve it yet.
(I must also point out that while the situation described above was very common, it certainly does not reflect *all* the applicants from Chinese universities. There are some institutions in China that give a first-rate CS education and I have had the priviledge of working with some excellent graduate students coming out of those programs. I'm just not yet convinced that this is representative of most cases.)
Anyone see the movie "Titanic"? Anyone remember how well it did at the box office?
To paraphrase: "Do not be to proud of this technological terror you've constructed... it is insignificant next to the power of the '13-year-old-girl' demographic".
>I'm a Sys Admin, and I firmly believe that what
>I do shouldn't be taught in college and
>university (and it isn't) but should be taught
>as an apprenticeship.
Now *that* is a good idea. A really good idea.
Like you said, the skills required to be a "grizzled database admin" or "unix guru" just can't be taught in the classroom... a great deal of it is problem solving, and the only way to get good at that is... practice. Even better when the practice includes guidance from an expert.
I'm intrigued by your ideas and would like to subscribe to your newsletter;)
Seriously though, I wonder if it isn't time to start some sort of organization that can lobby North American governments to move towards this sort of system.
I agree here; but I think we're sort of stuck with "Computer Science" for reasons of backward compatibility....
>Curious; as a tech school programmer,
>essentially a plumber, I have to ask: is all
>that you listed useful in any way for a
>programmer, in the long run?
Yes, some of it will be very useful. For example, suppose your boss comes to you and says:
"We're having serious problems with some of the junior coders writing programs that go into infinite loops and lock up our systems. I'd like you to write a program that will scan through a piece of code and determine whether or not that code ever goes in to an infinite loop".
Would you do it? If you have taken Theory of Computation you'd know that this is equivalent to the halting problem for Turing machines and thus is *impossible*. It seems bizzare, but there are actually a lot of things that we can't write programs to do, and furthermore, we can actually prove this fact.
So in some situations, I say "yes". In everyday coding? I'll be honest, most of it won't be used... but the one time you *do* need it, you'll be glad someone stopped you from attempting an impossible task, etc.
Another thing is that many companies working in mission-critical areas (autopilots, reactor control systems) have started requiring that contractors use formal methods for proving their programs correct. If you hire programmers that took semantics classes in college, they'll pretty much be ready to go... if they didn't... you are going to have to spend a *lot* of money on training.
If I were running a company, I'd want a mix of solid technical programmers and a few "theory guys". I certainly don't think every coder needs a CS degree (in fact, if programming is what you love, a CS degree is probably a invitation to pain and boredom).
>Sometimes I think the world just needs better
>plumbers. Or better plumbing
We do. Look how much a good plumber makes... its *not* an easy job and requires a very high degree of specialized knowledge and skill (much like computer programming). But still do need a few guys figuring out things like fluid dynamics so we can determine how much flow we can push through a pipe of a given length and diameter.
I whole-heartedly agree with you about the current situation though... far too many people are going into CS, when really what they want to do is computer programming. I see *so* many frustrated students in my office that have been mislead by ignorant councillors, etc... I'm not sure what the solution is though. If anyone has any ideas (or a few million $$ for a massive advertising campaign), I'd love to hear them.
>I've been programming since I was 12 (I'm
>currently 24) and have read hundreds of CS
>books. I think that I can easily complete a CS
>degree in 1 year.
"I've been using microwave ovens since I was 12 and have read hundreds of books on using microwave ovens. I think I can easily complete a degree in high-energy physics in one year."
Okay, thats pushing the analogy a little far, but you get the point. Programming is a trade skill, period. Computer science is NOT about programming. I'm sorry I'm reacting so violently to this, but you've hit a *very* sore spot for me.
I do computer science for a living and I am a mathematician by trade and training. Yes, I can program and enjoy doing so, but this is not what makes me a "Computer Scientist". I also enjoy tinkering with old sports cars and have a decent grasp of mechanics. I certainly don't consider myself a qualified automotive engineer though.
You claim to have read "hundreds of CS" books; but have you really? Is there a chance they were programming and technology books? If you want a taste of what real "computer science" looks like, I happily recommend reading the following:
- Computational Complexity by Papidimitriou
- Automata Theory and Languages by Hopcroft and Ullman
- Compilers... I can't remember the whole name but its got a big picture of a dragon on the front. If you ask anyone in the business about 'the dragon book' they'll know what you mean... by Aho, Sethi and Ullman (I think).
- The Russell and Norvig AI book
- Any book on lambda-calculus and recursive function theory (I can't think of a good introductory-level text at the moment). Even better if it introduces semantics too.
Those will give you a feel for some of the areas that are considered "Computer Science". The ability to program will get you through first year; after that, its more about math and.. gasp... thinking, than it is about whipping up code.
Again, sorry for the rant, but I think Edsgar Dykstra (a famous Computer Scientist) summed it up best when he said:
"Computer Science is as much about computers as Astronomy is about telescopes".
>perhaps the REAL source of artificial
>consciousness in the future won't be achieved by
>physical hardwiring of any complexity, but with
>some sort of "quantum ghost in the machine".
This is a very interesting proposition, and if you're truly interested in it, I would highly recommend reading some of the popular writings of Roger Penrose (The Emporer's new mind, etc.). One of his central theses is that 'mind' is a consequence of quantum effects.
Pesonally, I don't particularly agree with Penrose; but like it or not, I still find Penrose an excellent (and thought-provoking) read.
I have to say its wonderful to see TCS making an appearance on the front page of/.... warms my heart.
Anyway, after perusing the comments (most of them quite good), I thought I'd point out a couple of common misconceptions in previous posts:
NP != EXP. Yes, for problems in NP the best *known* algorithms require exponential time, but no one has proven this to always be the case. Maybe we're just bad at finding algorithms (not likely though). All we can say for sure is that problems in NP require polynomial time on a non-deterministic machine. Non-determinism being the key problem here (and what forces these algorithms into exponential time when you run them on deterministic machines).
NP != BQP. E.g. you can't solve NP-complete problems in polynomial time on a quantum computer. There are two types of quantum algorithms out there at the moment, Shor-flavour and Grover-flavour. Shor-flavour algorithms can solve some specific problems (notably factoring and discrete log) that are in the intersection of NP and co-NP but are *not* NP-complete. Grover-flavour algorithms can solve database searching problems in O(sqrt(n)) time, despite the classical firm lower bound being O(n).
Its most likely that BQP and NP are disjoint so attempting to compare them may be a rather fruitless effort.
If the hardness of NP-complete problems really causes you to lose sleep, look into the field of approximation algorithms. In many cases we can get a very good solution, just not the optimal one. We can even given an upper-bound on exactly how 'bad' the solution we find is compared to the optimal one.
As opposed to the situation now? How many patents do you think get filed by very small companies and individuals in a given year? Do you have ~$10,000 hanging around to throw at a patent lawyer?
I think the point being made on Slashdot is not that patents are inherently evil... just that the current system is somewhat broken because
a) its accessible only to those with the proper finacial resources
and
b) its pretty obvious that the patent clerks these days are *not* up to the standard set by Einstien...
I cut my teeth on HP3000s running MPE and, at least in my shop, *all* of our coding was done in SPL (systems programming language -- unique to the 3000 series). Even back then we knew that COBOL was a demonic construct.
Sure they're getting old, but I can't ever once recall one of these beasts failing us (unless someone was stupid enough to give a newbie coder 'Priveledged Mode' access.... ahh, the joys of that little PM flag in your security permissions).
I'm sad to see them go, and I think my duty here is clear: write an HP3000 emulator for PCs.... all the quirkiness of MPE with none of the fantastic 3000 hardware...
Slashdot Mirror
Let me get this straight... he wants hardware that will detect all possible programs that will copy digital media...
So, from a theoretical computer science point of view, he wants a Turing machine that will recognize all Turing machines that compute a fixed function f. That sounds remarkably like a problem that is equivalent (by reduction) to the halting problem for Turing machines... Oh, did we mention that the halting problem is unsolvable??
But hey, if *Mr. Valenti* says so, it *must* be possible. After all, everyone knows that you can simply legislate away fundamental laws of mathematics...
Whats next? Valenti proposing that we set Pi equal to 3.0 to simplify calculations?
>Is it just me, or do others wish that they could get the fan count down in their boxes?
Its not just you =) I recently went through the horrifying experience of upgrading my home machine from a PIII-450 to an Athlon XP (eXtra Pyrotechnics?).
On the upside, the money I invested in a new case and all those fans is more than offset by the savings I get since I no longer have to heat my apartment. There is the small matter of the power bills though...
>Why? Because we paid for it and own it.
The US "owns" the *whole* internet, and paid for it? Wow.. am I ignorant. I guess it makes sense though, since Al Gore 'invented' it.
As far as I know, ARPAnet and the TCP/IP protocol stack were invented by scientists working in America (but you might want to go back and check all their nationalities... you might find a few suprises!). By extension the early internet was mainly in the US. But hey, have you ever heard of JANet (Joint Academic Network)? You know the early packet-switching network that the BRITISH had?
Yup, true story. 'Back in the day', there was a single link between what you so quaintly refer to as 'the whole internet' (e.g. the US network that grew out of ARPAnet) and JANet. You had to telnet to a host out on the US east cost (name escapes me at the moment, it was a long time ago) and then enter the JANet address you wanted.
I seem to remember that JANet address were 'backwards' too (e.g. org.slashdot.www )
As far as I know, JAnet was designed and built pretty much autonomously by the British. I don't ever recall hearing about American's desigining it, or paying for it. In fact, I'm pretty sure it was based on X.25 too.
Anyway, the point is that several other countries had their own packet switched networks that eventually got hooked together to create the 'internet'. As the lowest common denominator its true that TCP/IP become the de facto standard, and yes this is an American invention.
But the honest truth is that the U.S. "paid for and own" the internet about as much as Al Gore invented pants.
>crist ever wonder why we have hospitaols that are way better than any crap you have?
Okay, I'm prepared to hear out your opinion if you can provide me with some actual *fact* to back up your ranting. Trying to make such a sweeping arguement with a single anecdote doesn't lend credibility to your argument.
Can you perhaps tell us what year of Med school you're in? You make several statements of opinion, but to be honest, you kinda have to show that you're an expert in the field before personal opinion becomes a valid arguing point.
Lets take a look at research productivity, shall we? Certainly this isn't the best measure of how good a health care system is, but its one thats very easy to quantify. And really, if there is enough time and money to do research, one assumes the patients are at least not suing for malpractice.
Grab a fistful of medical journals at your local university library. Count how many papers have first authors at Canadian institutions. Count how many at American institutions. Now factor in that our population is 1/10 the size of yours.
I'm not going to tell you what to think, I'll let you draw your own conclusions.
The hard reality is that if you're rich, there is no question US healthcare is better. The US is a capitalist county and if you work hard and are successfull, you can have *whatever* you want. Including the worlds best healthcare.
On the other hand Canada, which is really rather socialist, is a MUCH better place to be if you're poor. In fact, even if you're a street person, you'll receive the same level of care as a CEO worth millions of dollars (in theory).
If social responsibility is your idea of a "cluster-F**k", then I guess you have a point. If, on the other hand, you think that maybe as human beings we can evolve beyond a selfish ME FIRST attitude; I think you have to re-examine the facts.
Note that I'm not out and out supporting socialist healthcare policies here... I'm not convinced its the right way to go. India is a *very* socialist country and to be honest, I would not want to live there.
Anyway the point was that straight-out "x is better than y" comparisons of healthcare are really bogus.
US > CAN if you're rich; CAN > US if you're poor. If you're middle class, its not immediately apparent which is better.
The thing is, you'd never put an embargo on us because we supply you with such useful things as:
- Timber
- Grain
- Uranium
- Electricity (get your facts straight)
- Funny comedians
- oh yes, and a little thing called FRESH WATER
You really *do* need our natural resources, and for the record we have more oil here than in all of Texas. Its just that its in tar sands and really expensive to refine; but if we had no other choice, we could get along quite nicely.
I'd like to see you manage without water.
Amen Brother.
It was ludicrous claims like this about "the magical wonders of AI" that ended up with AI winter back in the 80's...
>I don't know what this guy's smoking.
A big pipe of ignorance... anyone who's taken a basic intro to AI course knows the crushing disillusion one experiences when one realizes AI is really just a bag of tricks.
Sophisticated and interesting tricks to be sure, but at the end of the day isn't it much more magical to believe the computer is really *thinking* about that game of chess... it sure takes the fun out of it once you learn about alpha-beta pruning, etc.
I wouldn't be surprised if this guy also thinks that David Copperfield really *can* teleport objects and levitate. I guess Clarke's law applies here, and to this particular individual, AI is a 'sufficiently advanced technology'... thus rendering it indistinguishable from magic and suitable for wild-eyed predictions.
>under current understood physics... Same goes for the rest.
I respectfully disagree... Godel's incompleteness theorem holds up even under alternative laws of physics. Though the whole P/NP thing doesn't (there was a paper on the xxx.lanl preprint servers a few years back showing that if quantum mechanics is non-linear then P=NP as a consequence. Good reading.)
I'm sure you can run Grover's algorithm in *simulated* O(sqrt(n)) time...
Did I mention that simulating a quantum system on a deterministic machine will require EXPONENTIAL time and space?
Go Perl.
You make some excellent points, but I disagree slightly with the following statement:
;) ) thoughts about Roots too...
;)
>The fact that there are corporations out there that
>are willing to give money to these athletes is
>enough for me to want to support those corporations
>so I don't mind knowing who the olympic sponsers
>are of the national team.
For those corporations "Olypmic Sponsorship" == "Tax-deductable advertising". Think about it. In your argument you mentioned Roots, so you were thinking about Roots clothing, and then I thought about Roots, and any other Canadian who read it thought about Roots too.
Even better... most of us had warm and fuzzy (no pun intended
On the other hand, it is nice that they sponsor our atheletes... altruistic or not. You made the excellent point that, regardless of intent, the atheletes still get sponsored.
Now heres something we can all agree on:
>I am totally surrounded by 24 hr olympics here on
>a sat. afternoon between my television, radio, and
>internet connection.. and comfy for 2 weeks of the
>hype.. I love it
Damn straight! GO Canada!
Thanks for dragging me out of my anti-capitalist cynacism
L.
>a Team Canada Patrick Roy jersey
If only... dear Patrick has decided not to play in the Olympics. Doesn't pay well enough I guess. =(
Our goalies are Belfour, Brodeur and Joseph.
Complete the following statement:
The modern Olympic games are about:
a) competition between amateur atheletes
b) the spirit of global friendship
c) Incredible advertising and product placement opportunities
d) making as much money as you can fit down your bloated gullet
e) all of the above
I guess I put option e) in there because I'm not a complete cynic yet; but lets face it... the games *are* commerical...
Its a good point, and you're correct that Canadian law *should* in theory have allowed this. If you want an example of what happens in practice though, check out the "icravetv.com" story and the resulting effect...
http://www.newsbytes.com/news/01/172941.html
Once again, it doesn't matter who's right... it matters who has more money. As long as you can buy... I mean *lobby* the CRTC and/or a good number of MPs, you can just get the offending law changed.
I think the following statement in the press release pretty much says it all:
>We perceive this advancement as a significant
>breakthrough to the historical limitations of
>digital communications as it was originally
>detailed by
>Dr. Claude Shannon in his treatise on Information
>Theory."
How about algorithmic information theory? Kolmogorov, Solomonov, Chaitin? The statement above indicates that the most recent word on compression is an old Bell Labs tech report by Claude Shannon... not to put Shannon down, that work *is* a landmark, but there has certainly been more work done since.
Try compressing the number Pi using Shannons theory... you can't do it. On the other hand, using Kolmogorov complexity, you can compress it quite nicely.
The fact that this statement appears in the press release seems to indicate a great deal of ignorance on the part of this corporations researchers. Part of any good research program is to familiarize yourself with previous work done in the field... and AIT is *not* some obscure backwater idea... there are several conferences on this topic every year and just about every CS graduate student has seen at least Kolmogorov complexity.
This is a pretty serious credibility robber. (Not to mention that from a mathematical standpoint, compressing totally random data is impossible under our current axioms... so if we *can* compress completely random data... its time for a new theory of the foundations of mathematics. At the risk of sounding dogmatic: do you *really* think some dot-com startup is capable of this?
Perhaps they are, but I'm going to need to see the proofs written up nice and formally before I run out and buy snake-oi... I mean *stock*.)
Its an interesting proposition, and I thought I'd chime in with my viewpoint because I think I have a (semi) qualified opinion here.
I work at a university which has a significant number of graduate students who received their undergraduate education from institutions in mainland China. Moreover, we literally receive hundreds of applications from such students every year. I served on the admissions committee for the graduate program last year and had the opportunity to review transcripts and CVs for many of these individuals.
I am of the opinion that what constitutes a CS education in China is radically different from what constitutes a CS education in North America and Europe. I saw several applicants who received degrees in computer science, yet the only courses I could find on their transcripts that were even vaguely related were "Visual Basic Programming I, II and III" and "Microsoft Access".
No theory courses, no programming languages, no algorithms, no data structures, no design, no software engineering. There were, however, mandatory 'Physical Education' and 'Marxist Economic Theory' courses. (I really feel for these students... I'm not sure how thrilled I would've been to have mandatory "gym class" in University!)
Unfortunately this is not a just a few cases; this was the norm (at least for the few hundred applications I've personally seen).
My overall impression then, is that these applicants probably are qualified to hack out VB code. In fact, their education seems to be setting them for trade jobs as coders. Unfortunately, whats lacking is any sort of rigourous training in basic computer *science* and software engineering. In the long run, one of three things can come of this: these individuals will educate themselves and become productive coders; someone else (i.e. their employers) will have to pay to educate them on the missing material; or, sadly, they simply will never be very effective software engineers. Since the people who will choose path #1 are few and far between, I'm not sure India has anything to worry about yet.
I think China needs to take a serious look at how they are educating their 'Computer Scientists' if they are serious about becoming a world software power. Its certainly within their grasp, I'm just not convinced they have the system in place to achieve it yet.
(I must also point out that while the situation described above was very common, it certainly does not reflect *all* the applicants from Chinese universities. There are some institutions in China that give a first-rate CS education and I have had the priviledge of working with some excellent graduate students coming out of those programs. I'm just not yet convinced that this is representative of most cases.)
Anyone see the movie "Titanic"? Anyone remember how well it did at the box office?
To paraphrase: "Do not be to proud of this technological terror you've constructed... it is insignificant next to the power of the '13-year-old-girl' demographic".
>I'm a Sys Admin, and I firmly believe that what
;)
>I do shouldn't be taught in college and
>university (and it isn't) but should be taught
>as an apprenticeship.
Now *that* is a good idea. A really good idea.
Like you said, the skills required to be a "grizzled database admin" or "unix guru" just can't be taught in the classroom... a great deal of it is problem solving, and the only way to get good at that is... practice. Even better when the practice includes guidance from an expert.
I'm intrigued by your ideas and would like to subscribe to your newsletter
Seriously though, I wonder if it isn't time to start some sort of organization that can lobby North American governments to move towards this sort of system.
>Mayhap it should be called something else,then.
I agree here; but I think we're sort of stuck with "Computer Science" for reasons of backward compatibility....
>Curious; as a tech school programmer,
>essentially a plumber, I have to ask: is all
>that you listed useful in any way for a
>programmer, in the long run?
Yes, some of it will be very useful. For example, suppose your boss comes to you and says:
"We're having serious problems with some of the junior coders writing programs that go into infinite loops and lock up our systems. I'd like you to write a program that will scan through a piece of code and determine whether or not that code ever goes in to an infinite loop".
Would you do it? If you have taken Theory of Computation you'd know that this is equivalent to the halting problem for Turing machines and thus is *impossible*. It seems bizzare, but there are actually a lot of things that we can't write programs to do, and furthermore, we can actually prove this fact.
So in some situations, I say "yes". In everyday coding? I'll be honest, most of it won't be used... but the one time you *do* need it, you'll be glad someone stopped you from attempting an impossible task, etc.
Another thing is that many companies working in mission-critical areas (autopilots, reactor control systems) have started requiring that contractors use formal methods for proving their programs correct. If you hire programmers that took semantics classes in college, they'll pretty much be ready to go... if they didn't... you are going to have to spend a *lot* of money on training.
If I were running a company, I'd want a mix of solid technical programmers and a few "theory guys". I certainly don't think every coder needs a CS degree (in fact, if programming is what you love, a CS degree is probably a invitation to pain and boredom).
>Sometimes I think the world just needs better
>plumbers. Or better plumbing
We do. Look how much a good plumber makes... its *not* an easy job and requires a very high degree of specialized knowledge and skill (much like computer programming). But still do need a few guys figuring out things like fluid dynamics so we can determine how much flow we can push through a pipe of a given length and diameter.
I whole-heartedly agree with you about the current situation though... far too many people are going into CS, when really what they want to do is computer programming. I see *so* many frustrated students in my office that have been mislead by ignorant councillors, etc... I'm not sure what the solution is though. If anyone has any ideas (or a few million $$ for a massive advertising campaign), I'd love to hear them.
>I've been programming since I was 12 (I'm
>currently 24) and have read hundreds of CS
>books. I think that I can easily complete a CS
>degree in 1 year.
"I've been using microwave ovens since I was 12 and have read hundreds of books on using microwave ovens. I think I can easily complete a degree in high-energy physics in one year."
Okay, thats pushing the analogy a little far, but you get the point. Programming is a trade skill, period. Computer science is NOT about programming. I'm sorry I'm reacting so violently to this, but you've hit a *very* sore spot for me.
I do computer science for a living and I am a mathematician by trade and training. Yes, I can program and enjoy doing so, but this is not what makes me a "Computer Scientist". I also enjoy tinkering with old sports cars and have a decent grasp of mechanics. I certainly don't consider myself a qualified automotive engineer though.
You claim to have read "hundreds of CS" books; but have you really? Is there a chance they were programming and technology books? If you want a taste of what real "computer science" looks like, I happily recommend reading the following:
- Computational Complexity by Papidimitriou
- Automata Theory and Languages by Hopcroft and Ullman
- Compilers... I can't remember the whole name but its got a big picture of a dragon on the front. If you ask anyone in the business about 'the dragon book' they'll know what you mean... by Aho, Sethi and Ullman (I think).
- The Russell and Norvig AI book
- Any book on lambda-calculus and recursive function theory (I can't think of a good introductory-level text at the moment). Even better if it introduces semantics too.
Those will give you a feel for some of the areas that are considered "Computer Science". The ability to program will get you through first year; after that, its more about math and.. gasp... thinking, than it is about whipping up code.
Again, sorry for the rant, but I think Edsgar Dykstra (a famous Computer Scientist) summed it up best when he said:
"Computer Science is as much about computers as Astronomy is about telescopes".
Truer words were never spoken.
>perhaps the REAL source of artificial
>consciousness in the future won't be achieved by
>physical hardwiring of any complexity, but with
>some sort of "quantum ghost in the machine".
This is a very interesting proposition, and if you're truly interested in it, I would highly recommend reading some of the popular writings of Roger Penrose (The Emporer's new mind, etc.). One of his central theses is that 'mind' is a consequence of quantum effects.
Pesonally, I don't particularly agree with Penrose; but like it or not, I still find Penrose an excellent (and thought-provoking) read.
I have to say its wonderful to see TCS making an appearance on the front page of /. ... warms my heart.
Anyway, after perusing the comments (most of them quite good), I thought I'd point out a couple of common misconceptions in previous posts:
NP != EXP. Yes, for problems in NP the best *known* algorithms require exponential time, but no one has proven this to always be the case. Maybe we're just bad at finding algorithms (not likely though). All we can say for sure is that problems in NP require polynomial time on a non-deterministic machine. Non-determinism being the key problem here (and what forces these algorithms into exponential time when you run them on deterministic machines).
NP != BQP. E.g. you can't solve NP-complete problems in polynomial time on a quantum computer. There are two types of quantum algorithms out there at the moment, Shor-flavour and Grover-flavour. Shor-flavour algorithms can solve some specific problems (notably factoring and discrete log) that are in the intersection of NP and co-NP but are *not* NP-complete. Grover-flavour algorithms can solve database searching problems in O(sqrt(n)) time, despite the classical firm lower bound being O(n).
Its most likely that BQP and NP are disjoint so attempting to compare them may be a rather fruitless effort.
If the hardness of NP-complete problems really causes you to lose sleep, look into the field of approximation algorithms. In many cases we can get a very good solution, just not the optimal one. We can even given an upper-bound on exactly how 'bad' the solution we find is compared to the optimal one.
Just my two cents worth.
L.
>the big companies will totally dominate,
As opposed to the situation now? How many patents do you think get filed by very small companies and individuals in a given year? Do you have ~$10,000 hanging around to throw at a patent lawyer?
I think the point being made on Slashdot is not that patents are inherently evil... just that the current system is somewhat broken because
a) its accessible only to those with the proper finacial resources
and
b) its pretty obvious that the patent clerks these days are *not* up to the standard set by Einstien...
>So it's all COBOL and EDI. Yuk
I cut my teeth on HP3000s running MPE and, at least in my shop, *all* of our coding was done in SPL (systems programming language -- unique to the 3000 series). Even back then we knew that COBOL was a demonic construct.
Sure they're getting old, but I can't ever once recall one of these beasts failing us (unless someone was stupid enough to give a newbie coder 'Priveledged Mode' access.... ahh, the joys of that little PM flag in your security permissions).
I'm sad to see them go, and I think my duty here is clear: write an HP3000 emulator for PCs.... all the quirkiness of MPE with none of the fantastic 3000 hardware...