Except it rarely works that way. I have some experience with both chess and Go, and have been the inexperienced player in that scenario more than once. I've yet to see even one single instance where it works like that from beginning to end.
The inexperienced player may pull one or two surprisingly good maneuvers out of sheer dumb luck, maybe even gain a temporary advantage out of those. But in the long run he'll fail to use and consolidate that advantage and the more experienced player will _bury_ him.
The chance to win a match by sheer dumb clueless doing something random that the other isn't expected is negligible because it just needs too many moves in a row where that happens. If the chance to make a surprisingly good and unexpected move is, say, 1 in 1000 (remember, it has to be not just good, but also some radical new strategy that noone tried before and the good player isn't expecting), then the chance to make two in a row is 1 in 1,000,000. And the chance to make 4 in a row is 1 in 1,000,000,000,000. Keeping up like that for a whole game is just not going to happen.
Plus, good players are good because they can adapt and use logic to different situations. He's not going to just give up and run in circles for the next half an hour just because you did one different move. He'll keep reacting and probing and you only need to get out of that lucky streak once or twice for him to fully use it against you.
Basically "beginner's luck" is a myth. It's a crap excuse by people who aren't as good as they think, to not admit that they played badly. Or that maybe they let you win. But if they didn't, then that supposed beginner actually played pretty damn well.
And if Google's secret sauce is "beginner's luck", then maybe all it says is that the big "experienced" players are the ones playing badly in that space. Maybe it's not Google who's clueless there.
BTW, I might point out that people like you used to discount quantum theory. "The world around me doesn't work that way, so it can't possibly be true!" they would say. Guess who was right?
"They laughed at Einstein. They laughed at the Wright Brothers. But they also laughed at Bozo the Clown." - Carl Sagan
Did I mention fallacies? I think I did. In this case the Association Fallacy (sharing one quality or in this case one group of enemies with quantum physics doesn't actually prove this one true too), salted with a bit of Ad Hominem ("people like you...").
Here's a fun reality check for you: quantum theory actually had some actual experimental data to show, and made some easily falsifiable claims. (Falsifiable meaning it could have been easily proven false, if it were false.) And in the meantime it's shown a mountain of evidence and made possible practical stuff such as LASER or the Zener diode. _That_ is what makes it true, not sophistry like who laughed at it or who didn't. That's how science works.
Unfortunately that idea is lost on some CS clowns who think that just arguing personal beliefs, or occasionally preferences, is some form of science. And that expertise in some unrelated domain makes them automatically right. ("I'm the compiler guru, therefore I can tell you which methodology is more productive." Or "I'm the data structures guru, therefore I can tell you what's needed for intelligence.") No, it isn't. Show me the actual experimental proof, and then we'll talk. Until then, it's no better than the work of all those who wrote from positions of authority about what's needed to transmute lead into gold.
And now you claim to know how the brain works. Are you really *that* arrogant? Yikes...
Believe it or not, _some_ things about it are known by now. There are more psychologists, neurologists, anthropologists and so on, all the way to stage "magicians", who've studied those various aspects, than CS clowns busy postulating they just know what's needed for intelligence. And the fun part is, the former group actually has some falsifiable data to show for their efforts.
At any rate, it doesn't matter if I'm arrogant (yes, I am) or what you choose to be incredulous about. Show me the experimental data and then we'll talk. That's how real science works.
Otherwise it ends up an "argument ouf ignorance" fallacy. When you've postulated that compression is _needed_ for intelligence, you've made a claim about the brain too. So it ends up basically, "We don't know how the brain on the whole works, so it must be true that it needed data compression to achieve intelligence." Which is a textbook example of that fallacy.
And the fallacy you've committed is decrying a theory without even reading about it, let alone understanding it. Tell me, which is worse?
Your using a fallacy yet _again_? Dunno about "worse", but it looks pretty bad. What you assume about me still hasn't proved or disproved the theory you're busy defending. What _would_ even start to prove it would be at least _one_ example of an AI where pure data compression was essential to its workings.
My point is that, with absolutely no basis in actual fact, you've decided to disregard what is, on the surface, a reasonably valid theory (to me, anyway), based on your personal "feelings".
Looking reasonable valid on the surface can be very mis-leading. Alchemy looked reasonably valid on the surface. Communism looked reasonably valid on the surface. Inteligent Design looks reasonably valid on the surface for a _lot_ of people. Etc.
Unfortunately, again, that's not how science works. Science works more along the lines of "show me the experiment". Show me that data compression is even useful at all in a working AI, before making broad claims that it's _needed_ for intelligence. It's that simple.
So, by the same logic, if you're not holding a Ph.D. in Economics, you're not allowed to say that an apology of Communism is bogus, right? They had plenty of Ph.D.'s in Moscow writing books about how great that system is. You wouldn't go and attack the work of people more educated than you out of sheer ignorance and stupidity, right?
Here's a fun concept for you: appeal to false authority. Because that's the fallacy you're committing there.
Show me someone who's actually built a working AI, and then we'll talk. That's a real authority I'm willing to accept any time. Anything else is, for the scope of discussing what's _needed_ for AI, just an unproven hypothesis. No more, no less. Even if their other maths is good, the supposition that's indeed needed for inteligent behaviour is no more than an unproven hypothesis pulled out of the ass. Arguing that since someone is an expert in the first, they must be automatically right about the second, is no more and no less than an appeal to false authority. It's like arguing that since Newton was good at physics, he _must_ have also be right about alchemy. (Which is another thing he studied, btw.) It's just that stupid.
So seeing that noone actually produced anything that can take a RL information stream and do anything even remotely intelligent with it, I'm not sure what makes them irrefutable authorities on the subject. Where _is_ the proof-of-concept program that uses data compression to actually extract useful knowledge out of Wikipedia and act intelligently about it? No, seriously. Anything else is just taking a guess. So exactly what there says that you should automatically stop thinking for yourself and take his unproven ideas for absolute truth?
In the meantime, however, we have this thing called a human brains that actually exhibits intelligence. And we're starting to understand some things that it does. Such as that, as I've mentioned, it does _not_ work by simply compressing the whole stream, but by extracting the relevant bits and utterly ignoring the rest.
We have _millenia_ of proof that if you're interested enough in the magician's left hand, you completely miss what his right is doing, or that if you focus on the miracle horse doing maths, you miss the owner giving it hints when to stop tapping. Heck, proofs that if they get your attention on something in the foreground, you'll completely miss the guy in a gorilla suit jumping up and down in the back, even made it onto mainstream TV.
So between some unproven hypotheses, and something that actually works that way, I don't know about you, but I'll go with what works here and now. I'll go with how the brain actually seems to work.
Here's some free, complimentary clue for you: "All that glitters is not gold." Just because someone wrote a ton of maths about compression, doesn't mean it's anywhere near applicable to building an AI.
And in fact, noone actually built yet anything even vaguely resembling intelligence yet, so maybe, just maybe, all those hypotheses are missing some vital point. Maybe, just maybe, just because you've compressed something to the minimum number of bits, doesn't mean jack squat about being able to use it effectively in real time. Maybe, just maybe, just because you've run arithmetic compression over a text, doesn't mean jack squat about actually being able to parse and navigate the semantic relationships in it. You just have a compression program and need some hours just to get your stream of bytes back, and be no wiser as to how to act intelligently on it.
In the meantime, however, we have a ton of psychologists, neurologists and, don't laugh, even stage magicians who've explored how the brain does work. You know, the only actual thing that actually exhibits some intelligence. We've worked out a lot of odds and ends about what happens there. We have some idea of what it filters out and when, how many seconds do some of the buffers hold the data, and so on.
So maybe you can take some hint from there. Maybe the information you need to extract and act on is really something like "[car][incoming]" instead of ivory-tower exercises like calculating the minimum number of bits you can compress the incoming stream to. That's the hard part: extracting the useful information there.
Or do you need something with lots of formulae? Read something about AI, lemming. There are a lot of people who've worked on various inference machines and such, and, surprise, none of them involved compression algorithms.
So, you know, speaking of "ubtracting out the sound of your own voice droning on at the drop of a pin about subjects you haven't made an effort to fully appreciate", maybe you can take your own advice there. Just because you get a boner about some maths, doesn't mean jack squat about applying it to a practical AI problem. Maybe start using your own head about how you'd actually act on that information instead. Just a thought, you know.
I can think of lots of stuff which would realistically count towards intelligence, but "compress knowledge" is the kind of thing that just sounds unbelievably stupid. In fact, it sounds like the kind of idiotic publicity stunt coming from a marketting guy who never heard of AI or programming.
The things you'd realistically need, and probably use every day unconsciously, include:
- good indexing: Just having the information somewhere in your head, well compressed is pointless if you can't quickly find it.
E.g., when you see a faucet, you need to immediately access such information as "turning it this way makes water come out, turning it the other way stops the water from coming out." You need to start from "faucet" and basically search for the information you need relating to "faucet". That's an easy example, but even trivial daily activities like driving your car often involve navigating a whole graph of associations, and if, say, a kid jumped in front of your car, you better not need hours to uncompress the data and find a solution.
- good "tokenization" so to speak. Think every word or concept being not a sequence of letters, but a pointer to that very concept. Think Wikipedia's hyperlinks. It may resemble compression on a superficial level, but that's like saying that a bird resembles the space shuttle on a superficial level. The goal here isn't just to minimize size by agnostically matching any chunk of text that came before, but to create a graph of interdependencies that you can navigate to find a solution.
E.g., again, if you started with "faucet" and went through "turn it clockwise", you need to imediately be able to reference "turn" and "clockwise" without sitting and sifting through a zip archive. And for that you need the real tokens, not any old chunk of matching text for compression reasons. A match for "turn" is good. A match for "rn i" between "turn it clockwise" and a previous occurence of "a barn in the backyard", as a compression algorithm would match, is utterly useless.
- good filtering. Yes, the brain is sorta lossy at all stages, because it mercilessly filters out everything that doesn't look relevant or in the focus of attention. E.g., when looking at the street, you might just see "it's a car", because the details about it don't look relevant at the moment. E.g., when focusing on the car for more details, you might completely miss the guy in a pink gorilla suit doing cartwheels in the background, because it looks irrelevant to your current focus of attention.
This is more important than it sounds, and is the _real_ compression for the most part when processing that data. The brain isn't just compressing the whole scene to save bandwidth, it's really mercilessly clipping 90% of it out, tokenizing the rest, and again clipping out most of the details about those tokens as long as they're not explicitly required. It's not just compressing the image of a car coming at you, it's transforming it into the "tokens" (so to speak) "[car] [incoming]". Anything else from that scene (e.g., the guy in a pink gorilla suit across the road) isn't compressed, it's filtered out, and so are the details about those tokens. (E.g., car model, colour, registration number, etc.) You can choose to pay attention to some of them (e.g., look at the registration number or the driver's face), and that will change the filtering criteria to give you those details at the expense of something else.
Even what you mention ("while my brain is good at remember the gist of knowlege, but really bad at losslessly recalling it.") is just an example of such filtering. A lot of the details were just filtered out before you even memorized that (e.g., you might have gotten the idea that a giraffe has a long neck, but not its colour, because it didn't seem important at the moment), an a lot just weren't stored or indexed or given a high priority. If it didn't look useful to remember what colour a giraffe is, it didn't get stored. Again, it's filtering, not dumb stream compression.
And so on.
And missing all that in favour of a "let's see who compresses the best" stunt seems just stupid and clueless.
Reminds me of when some other such database (which, as it turns out, powered our corporate firewall) flagged a big chunk of MS's website as drug-related IIRC.
If someone forced a collect call on me with no more warning than "someone is connecting your call", I _might_ call free() or delete() on their private parts.
And this, kids, is why you should use getters and (if needed) setters instead of making an Orkut-inspired design where everyone is "friends" with everyone they ever heard of. If anyone needs your private parts, they can ask nicely and you can give it to them in a civilized manner.
There are plenty of people on Slashdot who hold contradictory opinions in order to blame MS in all possible situations even if you aren't one of them.
Sadly, this isn't slashdot-speciffic, but a general tendency of most people. People tend to not juggle too many variables about someone or something. ("John is a great coder, but he's an asshole, he's too introverted to manage a team well, and he's an average driver, and...") Even if they acknowledge such things as different factors, the subconscious tendency is to take the overall impression and apply it to everything. So if they like John a lot, they'll tend to view everything about him in that positive light. ("John is a great coder, and refreshingly frank and honest in dealing with people, even if that annoys some, has a refreshing hands-off approach to managing his team, and drives great, and...") If they dislike John a lot, they'll view everything about him in that negative light. ("John is an asshole who couldn't code his way out of a brown paper bag, is the most clueless PHB ever, and drives like a retard on hard drugs.")
People who look good and are well spoken tend to get the promotion, and as at least one study showed, they don't end up in prison as often. If you're in the jury and like that guy a lot, your natural tendency is to extrapolate that positive impression to everything about him. Such a nice guy _must_ be a honest and hard working guy, and probably got framed for that crime. If he's an ugly guy with a bad accent, well, the opposite applies. The motherfucker must be a no-good bum, and thank goodness we can put him behind bars.
The same applies to products. If they were made on the same chasis, with the same engine, and handled exactly the same, you'd probably still be inclined to say that a BMW handles better than Homer Simpson's dream car.
The same applies to companies. If you like a company a lot, then their products must also be the best of the best, their salesmen are honest, their employees are the most brilliant guys that ever walked the Earth, their patents are breakthroughs comparable to inventing the wheel, the waste they dump in the river is just pure water, their factories don't cause global warming, and their lawsuits are right and justified. That's why marketting and PR departments try to create a good image for the company as a whole.
And conversely, the same applies if you really don't like a company. Then their products by definition suck, their salesmen are backstabbing snake-oil salesmen, their employees are the rejects of vendor-machine-refilling school, etc.
It's not trolling, it's just extrapolating the general impression to each of the components. If someone would get a B grade on the whole, the tendency is to act as if every single component is the same B grade as the average.
It's not logical, but that's how humans work.
So some people apply that to MS. Whop-de-do. Humans acting like they're human on Slashdot. Who would have guessed?
"Adequately justified" is often just a matter of creating enough scary hype. See absynthe which is now legal again, after discovering that _no_ substance in it except the alcohol has an effect on the brain whatsoever. Yet a massive media campaign at one point presented it as a dangerous drug that turns ordinary people into crazed bloodthirsty serial killers and such bullshit. And enough hype and cerrypicking conjectures eventually were "adequate justification" for its prohibition.
Marijuana just had the mis-fortune of being hemp, and paper-from-hemp was at the time a threat to someone's paper-from-wood industry. (Yes, you've read it right. It had really nothing to do with people smoking it.) Unfortunately that someone also owned a media empire, and he used it to immediately unleash a barrage of fearmongering and bogus stories about crazed Mexicans going on murderous sprees after getting high on hemp. Even the name "Marijuana" started there, to make it further sound like a Mexican thing.
(Using the racism and nationalism card is a pretty common theme with most prohibitions and scare campaigns.)
At some point that was "adequate justification."
After that, and the ever increasing evidence that it doesn't actually _do_ any of that, it gradually seems to have backpedalled into:
- "yeah, but it's a gateways drug" (which not only was never proved, but also never answered the question "then why doesn't alcohol act as a gateway drug, if people are somehow made to progress to harder stuff?") and
- "yeah, but people end up stealing and mugging to support their drug addiction" (except it wouldn't happen if it was legal and reasonably cheap, like, say, tobacco is), and
- "yeah, but it's trafficked by mobsters and other criminals you wouldn't want in your neighbourhood" (who'd be out of job if it were legal), and
- "yeah, but once people start breaking the law to smoke pot, God knows what other crimes follow naturally." (The "gateway" argument all over again, and again missing the point that it wouldn't happen if it wasn't illegal to start with.)
More importantly it misses the point that Holland is an uncomfortable example of a country where it is still legal, and nothing bad happened. They're not overrun by gangs of drug-crazed psychos, and in fact (like most of continental western Europe) their criminality is actually quite low. It's not a gateway to harder drugs, either. Even assuming the "gateway" arguments wasn't bogus (it never was proved), if you can get your hemp legally, you're not that tempted to switch to something which can get you arrested instead.
But, eh, if you keep the hype campaign going, Jack Average won't think much further than what the idiot box tells him to think.
Think of it this way: look at how many bits the human DNA has. Each nucleotid pair is one of 4 possible values, so 2 bits. A human has about 3 billion pairs. So on the whole about 750 megabytes or so.
I dunno about you, but I'm thoroughly impressed. _You_ try programming a human in 750 MB, and then you can criticize. I'm talking not only the brain (which is a feat by itself), but also the whole organism there, including immune system, self-healing, metabolism, etc.
I don't even know if it's unstructured. What we have here is some people trying to use that stuff without knowing what it really does or how it was supposed to be used. It's sorta like watching the client PHB trying to just drag and drop a button on a form himself, and then wondering why it doesn't work like he just assumed it would. ("Hey, I dropped a "Save" button there and it doesn't actually save when I click it!") You can't really blame it on the original programmer if it doesn't work that way.
So if I were mean, I'd up the ante and also say "and program it so it also works when some clueless guy later tries to use a human cell in a mouse, in a way that was never in the specs." But let's be generous and skip that. Just program a human in 750 MB, no matter how.
What _our_ engineers (and I'm one, so I'm allowed to criticize) manage today in 750 MB is a stupid text editor or a spreadsheet. We're past even just structured programming. Now we pack everything in layers upon layers of frameworks, EJBs, factories, decorators, managers (the pattern, not the PHB), events, SOAP, JMS, etc, just because it's _fashionable_ to have one more buzzword on the resume. And, oh, if it's Java, let's add add a layer of introspection too, because it's become soooo unfashionable to just write "myObject.getID()" instead of cracking the class open the class at runtime and reaching in its internals.
Let me stress again: this crap doesn't even have to do with OOP or with structured programming any more, and in fact sometimes _prevents_ proper OOP. Project after project I run into crap designs where you _can't_ use OOP, e.g. define a simple subclass of something, because for example there's an EJB layer in the way and the other end wouldn't know how to deserialize your modified objects. Or because all the data objects are generated from some XML definitions -- and I don't just mean the stuff that'll get persisted in the database, but really internal data objects -- just because someone thought it's _cooler_ to write an XML and run a third-party generator than to just write the fucking member definitions and ask Eclipse to make getters and setters for them. And as a result you can't even attach the relevant methods to them, like you learned in the OOP classes, like attaching a "findChild()" method where it belongs in the tree node, because it would get overwritten in the next build when those objects are generated from XML.
Engineering used to be about having a problem and designing the best solution to it. E.g., you have a river to cross, and you consider whether it's best to build a bridge, or a tunnel, or a ferry, and pick the simplest and cheapest solution that solves the problem. Now we're at the stage where we want to have a "suspension bridge" buzzword on the resume, and we'll cheerfully dig a canal just to have something to build that bridge over. Or detour the road through 3 states to the side so we can find a gorge to build the bridge over. Must have that precious buzzword even if it kills the project. That's not engineering, that's marketting and playing.
Anyway, at this rate we'll soon need a 750 MB framework just to pass the parameters around.
It's just a funny article - don't take it too seriously:)
Except it's not even actually funny. There are funny reviews, there are good no-holds-barred "look at all that's bad with these games" reviews (e.g., on Something Awful), and both are good and refreshing in a world of bending over to EA's marketting money. But this is neither. This is one guy ranting incoherently, while trying to sound like a hip wisecracking smartass, but failing miserably.
It's just trying to construct some convoluted mixed-metaphor insults that go on for a paragraph. And more often than not are of the kind of "if you dare disaggree with me, then here's in how many ways you're a retard huffing spray can and probably also have fun banging your had against the inside of a discarded refrigerator." I've read/. trolls that were funnier to read, and could make a point without preemptively insulting anyone who might disaggree.
Basically what I'm saying is, insult the game if you must, but if you can't make a point without ad-hominems then you probably don't have a point to start with.
They say that if you try too hard to make an impression, that's the impression you'll make. And this is proof of it, if anything. The guy is trying too hard. I'd rather have had him do a professional review of why a game is bad, and keep the wisecracks short and to the point. A one-liner to make the point is good. A two-paragraph convoluted argument about why the reader is gay and a retard to boot, out of a maybe 8 paragraph "review", is just a piss-poor substitute for actual review content.
And whatever content it had, it's mostly picking on some 10 to 15 year old games and whining that they don't have the graphics/content/controls of a 2003 game.
Some of the screenshots there are from games from the 80's and, frankly, don't look any worse than any other game released in the 80's. Sure, they look bad by today's 3D pixel-shaded bump-mapped ragdoll-animated standards, but back then pixelated midgets and 4-colour backgrounds were pretty much the norm. Ditto for the animations: you had maybe 1 or 2 attacks for yourself, and maybe 1 or 2 for the enemies, and that was that. Cloned enemies repeatedly kicking or throwing purses at you were the norm. You just didn't have the memory to store detailed life-like backgrounds and a gazillion life-like animations.
And yes, occasionally punches and kicks went right through a character's chest, because you didn't have 3d ragdolls to stop a kick exactly where it touches the enemy. If your sprite of a character had his leg kicking at a full 3 ft distance and the enemy was 2 ft away, yeah, your foot would go right through them.
Some seem to be from the age when everyone discovered they have just enough RAM to use scanned photos for the characters. In many ways it was a worse age than the aforementioned pixelated-midgets age. At any rate, it still had the problems of the previous age, with an uncanny valley factor thrown in. Now you had realistically scanned enemies acting like stiff cardboard cutouts. Or in some cases you just had blurry contrast-less scanned midgets instead of the previous pixelated ones with good contrast.
Don't get me wrong, I'm not saying they were the great games of their era or anything. But it seems kind of silly to pick on the graphics of an 80's game just because they're 80's kind of graphics.
This is another cost. And said PHBs can only do so much non-work before they destroy the company they are part of.
Very true, but I do know someone who was left for a whole year to drive a company into the ground before he was fired. Not just non-work, but he caused all programmers and designers to quit together, he alienated all the business partners that company ever had, etc. (He was a complete asshole to employees, customers _and_ the companies he subcontracted to.) You'd think it would be obvious that something is wrong by the time he has no more employees he can make the programs with, and has to subcontract everything, but it obviously wasn't obvious to his superiors.
So, again, just because something or someone is counter-productive, it doesn't mean it won't happen. Sad, but true.
"Zero risk"? These jobs aren't zero risk, not remotely. A data entry clerk with a drug habit is a great way for credit card numbers and other valuable bits of information to get out.
Not every data entry typist has access to the credit card numbers. (And if all do, then there's something wrong with your database and applications.) There are jobs as boring as just correcting the format of entered telephone numbers, for example.
Plus, people with a big drug habit aren't reliable.
This is actually the mis-conception that bothers me the most, and sadly it's been hammered into a lot of otherwise intelligent people's heads by decades of propaganda and lies. Everyone thinks that if someone does a joint or two a day, they're inherently becoming slackers, thieves, maybe murderers, and they'll probably switch to harder drugs soon too. And that's just not true.
I don't do drugs myself (I figure I'm already ruining my lungs enough with tobacco anyway, plus I like to stay Lawful Good), but I've worked with people who smoked pot. They were no different from any other human. In fact, I've worked for a couple of years with two of them before finding out they did pot, and I didn't suspect a thing.
No, they weren't unreliable. No, they didn't sell company data. (And the two I've mentioned above were the DBAs, so they had access to literally everything in our databases, if they wanted to.) No, they didn't steal anything. No, it wasn't a gateway to hard drugs.
If anything, in retrospect they were calmer and more professional than some other coleagues, although I don't know if that's due to pot or that was just their personality type. They just did their job in situations where I'd have been at someone's throat for doing something stupid. But again, it might just be their normal personality for all I know.
And frankly it bothers me to see such people being villified and victimized. It bothers even more to know that a chunk of my tax dollars goes into victimizing such people.
Heh. This is slowly getting funny. One conspiracy theorist was interesting, but two thinking they can "prove" their point with repeated appeal-to-hidden-motives fallacies and conspiracy theories... now that's funny.
Folks, if you want to argue physics, go read a physics book first. Wishful thinking and appeal-to-hidden-motives phalacies are _not_ how science works. Show me some real physics data, not rants about GM and Exxon conspiracies, not copious hand-waving, and not verbal fallacies. Those do not quite a scientific proof make. I don't freakin' care what GM or Exxon think. Try reading at least a high-school physics book and using your own head for a change.
And give me a break with the wiring size from battery weight bullshit. My BMW 328iS has the battery in the trunk for a 50/50 weight distribution, the cable don't mean shit.
Again, let me enlighten you: an electric motor is made of coils through which current passes. Big thick coils consisting of layers upon layers upon layers of wire, like big spools of thread, on a ferrite core. That creates the magnetic field that creates the torque. (Actually, lemme play it safe and translate "torque" for you too: it's what makes it turn.)
_That_ is the wire I'm talking about. That mile of wire in those coils. Not the battery cables.
So, yes, the battery cable doesn't mean shit. Too bad your knowledge of electric motors means just about as much too.
You had me going there for a while, but this is just totally wrong. And it's blatantly wrong in favor of your viewpoint, which makes me re-read everything else you've written, filtering for strong bias.
Ah, way to start a good old fallacy. You can't actually make your point without attacking the other's supposed motives, can you? Well, anyway, let's ignore that and have a look, shall we?
Higher acceleration doesn't mean more energy consumed. The amount of energy to get from one speed to another is exactly the same (ignoring friction) regardless of acceleration.
"Ignoring friction" is good and fine in an ideal world, but in the RL it isn't. Heck, why stop at friction? Let's ignore all resistance for that matter, and then you don't even need an engine to keep going.
Different motors have optimal efficiency at different accelerations, so it might actually be more efficient to accelerate faster.
Heh. Nope. Funny, but just nope. An electric motor provides pretty much constant torque at any RPM, and regardless of how that RPM varies. Pretty much the torque (hence acceleration) limit for a given motor is how much current you can run through it. If anything, running more current through it just means more losses, so you're really saving nothing by flooring the "gas pedal" instead of starting slowly. Au contraire.
If you have regenerative braking, it may not matter to have more mass (no pun intended). You invest more energy into speeding up, but you get it all back (subject to inefficiencies in the regeneration).
Ah, again ignoring RL. _If_ you had a perfect system, where nothing is lost, the world would be such a wonderful place indeed. Unfortunately RL doesn't quite work that way.
Plus, you'd save exactly as much there with a hybrid car using an ultra-capacitor, so it's hardly anything specific to going all electric.
And, of course, you conveniently ignore the fact that a regular car has even more weight to accelerate (with no regenerative braking) than the electric car. So what if the batteries weigh something? A gasoline car lugs that big ol' combustion engine around, and it weighs something, too.
Heh. How about giving me some numbers, then? That big ol' combustion engine _and_ the fuel tank still weighs a fraction of what the batteries would weigh. Let me do some maths for you:
Gasoline stores some 14kWH/kg. About a quarter of it is actually useful, and about 75% are losses of the that big ol' combustion engine. So if you filled your tank with, say, 20 kilos worth of gasoline (about 25 litres, or about 8 gallons), you'd have about 280 kWH stored there, out of which some 70 actually help move your car.
For the best Li ion batteries available nowadays, the density is 120 WH/kg, or 0.120 kWH/kg. (For lead-acid it's 0.04 kWH/kg. For ultra-capacitors it's 0.003 to 0.005.)
Let's also assume that you have an ideal system where the electrical car is 100% efficient. It isn't, but let's pretend for a bit, shall we? So for this ideal car we'd need to store only 70 kWH to be equivalent to the above quantity of gasoline. Right?
So you'd need 70 / 0.12 = 583 kilos worth of batteries. I don't know what car you drive, but most internal combustion engines are lighter than that.
To put things into perspective, for an Audi A8 4.2 TDI Quattro the engine weighs 255 kg. And here we're talking a big freakin' 4.2 litre 8-cylinder dual-turbo Diesel engine. The 4.2 litre 8-cylinder gasoline version weighs in at 190 kg. The BMW M54 engine, a big ol' 3.0 litre 6-cylinder engine, weighs 170 kg. A Ferrarri F50's engine weighs 198 kg, and that's a big ol' V12. But these are powerful sports engines. Your average small commuter car will have a much smaller engine.
But so far we've been making the comparison based on the flawed assumption that the electric engine weighs n
Which is still rather irrelevant if you use a coal motor (ok, steam turbine) to produce that electricity in the first place. No matter how efficient you make that electric motor, the efficiency of it all is still basically the product of that _and_ the turbine that produced it in the first place.
Which, I do believe, was the GP's point.
Electricity can come from non-polluting sources.
Like... nuclear power plants, for instance? Because everything else either pollutes (e.g., coal plants) or just doesn't quite scale like some people seem to think. If you think you could just run whole countries _and_ their cars off wind turbines, solar energy and hydro plants, I want to know what planet you live on.
I.e., again, you actually just make the GP's point. Either you charge that car from a nuclear plant, or you've just moved pollution somewhere else. That was the whole point. Being both anti-nuclear and anti-pollution is just an unattainable proposition, short of returning to the stone age.
The cost of electricity hasn't risen 300% in six years.
Unfortunately, a large part of the reason there is nuclear energy, so again you just make the GP's point.
Another reason is that electricity consumption didn't rise that much either. Most of the manufacturing industry actually migrated _out_ of the USA, so industrial consumption actually dropped, helping somewhat offset the rises in other parts. And home consumption didn't rise that much either. And to stay "on topic" you don't yet have every Jack Redneck driving an electric pickup-truck and Jane Soccermom driving an electric SUV, either.
Pollution from a few sources is more easily managed and disperses less than from millions of ground level sources.
Heh. If anyone knew how to make a filter that stops CO2, we wouldn't be having the whole global-warming hysteria in the first place. So _how_ are you going to manage that? No, seriously, I want to know.
Rest assured that burning coal at a power plant to produce power for an electric car releases would release _exactly_ the same quantity of CO2 into the atmosphere as burning it directly in your car... if distribution, batteries and electric motors had exactly 0% losses. As it is, it actually releases _more_. Ah, but a car runs on gasoline, not on coal, so it's already getting better for the cars.
Ditto for dispersion. How are you going to keep the CO2 from dispersing into the whole atmosphere? No, really.
So, again, you just make the GP's point. Either you charge that car from a nuclear plant, or no, we don't actually know how to manage CO2 pollution from factories either.
Electric cars are simpler mechanically, more reliable and easier to repair.
Which is irrelevant to the problem being discussed, so I won't get into that either.
Electric cars accelerate faster and can use regenerative braking.
Yes, it _can_ accelerate faster, but that's irrelevant to the pollution topic. Well, not quite: accelerate faster == use more power.
Yes, it can even get some of its energy back when braking, but that's losing sight of the fact that the electric car has to carry half a ton of batteries too. So it will actually expend extra power into moving all those batteries (doubly so if you want to move them faster), which means extra pollution at some power plant down the line. Unless, of course, you use nuclear power.
Existing range limitations can be overcome with improved battery chemistry.
Unfortunately that just boils down to "it could get better in the future" optimism. Yes, battery chemistry will probably eventually improve, but here and now it still sucks big fat hairy arse compared to gasoline. Here and now driving an electric c
They rely on big huge stars to make big fat elements, then explode spreading them all over the universe where the coalesce into planets like the Earth.
Actually, that's not entirely correct. No star we know produces elements heavier than iron and nickel, which aren't very radioactive. In fact, they're the most stable nuclei we know.
The thing is, anything lower than iron and nickel tends to release energy when fused into something heavier. Anything heavier than that needs to absorb energy to fuse into something even heavier, and conversely releases some energy when split.
So eventually the reaction stops at iron and nickel. Given intense photon bombardment in the star, most nickel actually disintegrates right back into smaller nuclei, not fuse further into heavier stuff. Iron pretty doesn't do anything whatsoever, and just stays iron.
The thing there is that as you move upwards, the energy and temperature requirements tend to be insane. For example for the next step up from fusing hydrogen into helium, it takes a red giant and temperatures of about 100 _million_ Kelvin to even fuse helium into carbon before blowing itself up.
And most stars either (A) stop short of even that and become a red dwarf, or (B) blow themselves up within seconds when they start fusing helium, because that's a very unstable reaction, whose rate increases with temperature, and temperature increases with fusion rate.
But at any rate, even if you had a star massive enough, you wouldn't get many nuclei past iron, or you wouldn't get them out of the star. By the moment a star got massive and hot enough to start fusing iron into something heavier, it would just rapidly lose heat in that reaction. It just can't explode that way, so at most you'd get a black hole in the end of it all.
So since you mention stars exploding... well, that's actually where the heavier elements come from. Supernovae don't just spread those heavier metals, they _create_ them. The iron, carbon, helium and whatever else was created will be smashed with tremendous amounts of energy and at insane temperatures, and a lot of it will fuse into heavier stuff. And since the star is already blowing up, they'll get spread all over the place.
Employee monitoring outside of the workplace, especially secret monitoring is expensive and frankly not productive. After all, what sort of employee will consent to that kind of thing? How would that affect morale? I can see legitimate if paranoid special cases where monitoring might be worthwhile, but in those situations they should be writing a pretty big check anyway.
1. Some forms of monitoring are actually dirt cheap.
To start with the obvious, spyware is pretty ubiquitous at some companies, and that includes company laptops. So then people take them home and use them too for IM, slashdot, VOIP, updating their "anonymous" blog, and whatnot, and you can see where that is going.
E.g., someone posted a while ago, in a thread about tele-commuting, about how he knew an employee wasn't really working at home because he looked on XBox live all the time and after a couple of weeks the employee had 5 achievements in Oblivion. (Never mind that Oblivion is a game which can be finished in a weekend if you just follow the main story, or in a week without telecommuting even if you do every single side-quest. And 5 achievements aren't really that much.) That's a form of surveillance.
Google can also be used as a cheap form of surveillance, because most people don't really try to be anonymous. Or can be identified by details they provide.
Cell phones can also be tracked, as proven by a recent article, but I didn't bookmark it. Basically a journalist used such a tracking service on his girlfriend's phone. It asked for confirmation once at the start, and from there it was basically in stealth mode. In that case it was with her knowledge, for research purposes, but you can see how that can happen without knowledge too, if you have access to a "logged-in" phone for a couple of minutes. Company cell phones are a prime example: they can be subscribed to tracking before you even get the damn thing.
2. The line of reasoning that something won't happen because it's not making any money (or preventing losses) for the company is flawed too, and assuming that humans on the whole only do perfectly rational stuff supported by solid logic and numbers. That's false. Humans do a lot more for emotional reasons than for anything even vaguely resembling cold logic supported by facts.
Some PHBs (A) have nothing better to do with their time (even doing lunch and painting powerpoint foils only takes so much time), and (B) are complete control freaks. They don't do it because it actually helps the company in any form or shape, but just to feel in control of something they actually don't really know how to manage.
Even HP's case, if you look at it, is really no more than some control-freak exercise. If you look at the "leaks" they were investigating, the grand acts of treason to the press so to speak, the mind boggles. One executive had unauthorizedly told the press that he's tired after a long board meeting. Or that HP hopes to sell more of their Opteron servers in the future. (Well, of course. Is their any company who actually hopes to sell less and lose market share?) It's benign, uninformative and bloody useless small talk, not any actual company secrets.
But someone was chuffed that a director dared talk to the press at all, even such uninformative small-talk, without their royal seal of approval. I.e., a control freak. That's really how that espionage and stalking affair got started.
3. Even when logic and facts are involved, a lot more often than not, the goals are PR, looking good, etc, not "is it making the company money." You can see it from company policies and politics to PHB's more concerned with maintaining an illusion to their superiors than with managing what they're supposed to manage. Whole man-years get spent on just seeming to do something about a problem, instead of just fixing it.
Or to take your example with drug testing, the thing is: people aren't testing only investors and board members. You know, people who could actua
See, before humans figured out how to smelt iron out of ore, there were weapons made out of meteorite iron. A certain number of meteorites are nearly pure iron, and better yet, some is even already alloyed with stronger metals. They were rare and more expensive than gold, but it was a weapon which could pierce right through a bronze cuirass, and was often credited with magical properties. Kings and nobles paid a small fortune for them.
Some of the myths around that kind of equipment persisted even after it was known how to just smelt iron ore. E.g., the celtic myths about cold iron against elves. The only iron that can be processed without heating from start to finish is, you guessed, a chunk of stuff that was weapon-grade iron from the start, not ore. That's more often than not a meteorite.
So other than maybe modern times and construction crews with bulldozers, you wouldn't just throw away such a rock if you found one. You'd sell it to a smith for a small fortune, and he'd make a weapon for a king and sell it for a bigger fortune.
Dunno about Apple, but if I were mass producing those things, I would _not_ build the thing empty, connect it to a Mac by hand, transfer the stuff to it slowly via Firewire, etc. That kind of "let's connect a cable, launch this handy application and click here to transfer the files" is ok for a mom-and-pop shop, but when you're mass producing stuff you just want to shave the last penny off the manufacturing costs.
So the way it's done is you take the working prototype, make an image of its hard drive, and write that on every hard drive before it's even assembled into the iPods.
Think, basically, how your IT department doesn't come with a suicase full of install CDs for Windows, Word, etc, for each PC. They just make an image off one workstation and then install that on all others. Much faster.
Same thing here, only more automated.
So if that image was made from a HDD with the virus on it, the assembly line will mindlessly churn thousands of copies of that.
That's actually the problem. Your average PHB indeed doesn't know jack shit about the difference between ReiserFS or FAT, or between Java and Visual Basic. So he'll take that kind of decisions not based on their actual merits, but based on rumours, over-simplified half-truths they half-understood from some IT-for-managers ragazines, fashion, and what the nice MS/IBM/whatever salesman filled their head with during a round of golf.
I've seen people actually take such stupid decisions as "let's use a single-user database and just copy the database file on the department's file server", in that case MS Visual Fox Pro for a reason as stupid as "Visual Fox Pro is more visual than Java". Once the nice MS salesman showed them some dragging and dropping buttons around (and, as everyone knows, there's nothing else to programming an app than dragging and dropping the buttons on forms), any other considerations like concurrent access, transactions, available tools and libraries, etc, went right over their head.
So the danger is precisely that at some point a nice salesman shop drops by and goes "whoa, you guys run SuSE? Did you know they paid a convicted murderer to develop their filesystem? Every time you save your powerpoint presentations on that file server, you have an innocent's blood on your hands, not to mention all over your neatly formatted presentation. Now if you upgraded to Vista Super-Professional Snake-Oil Edition, you'd show your support for the Bill and Melinda Gates Charity and be _much_ more fashionable among your peers."
No. Just no. See, the world is full of idiots doing something stupid and pointless. Those are a dime a dozen, and comparing them to people like Columbus or Magellan is just insulting to the latter category. The ones who changed history, e.g., Columbus or Magellan, weren't retards doing publicity stunts, they were people who put some solid thought into what they were trying to do.
E.g., Columbus's calculations might have been wrong, but he started from solid evidence that the Earth must be round. The idea that travelling West can get you somewhere in the East was sound. If there hadn't been this unknown continent in the way, he would have actually arrived in Asia as he planned.
E.g., Magellan actually did even better. He set a goal, put some solid thought into it, and actually achieved it.
Do you understand? That's the difference between those and Yahoo's publicity stunt. In the comparison to Columbus and Magellan, Yahoo comes out just about comparable to the village idiot running around with pencils up its nose. Sure, it gets some attention and maybe even some money, but comparable to Columbus and Magellan it ain't.
That is, assuming Hanlon's Razor applies. ("Never attribute to malice that which is adequately explained by stupidity.") But there's also a chance that it's just plain old dishonest, and they don't actually give a shit about actually achieving anything except publicity. I.e., pretty much as ic Columbus had just stirred up a frenzy, got some money out of it, then went and got drunk in France with that money. Without even giving much of a damn about actually reaching the Indies. Hey, he did his publicity stunt, got his money, the rest isn't his problem.
IANAA(lien), but I'd quicker beleive an alien society could share visual or audio abilities with us than lingual ones. Decoding images seems more reasonable than having to learn morse code AND the language it is in.
If you gave an alien a photo, yes. But if you give them a signal encoded with some proprietary lossy compression? Heh.
The problem isn't the photo. The problem is that you get a stream of 1 and 0 and you have to figure how the fuck to even make heads or tails out of it. Before you'd get a photo to use your alien intelligence on, you'd have to start from scratch in, basically, cracking the encryption. Because that's what a JPEG is to someone who never heard of the format specs: as good as encrypted.
And you don't even have any known cleartext to aid you in breaking that code. What are those 1s and 0s trying to tell you? When you should stop trying new things on that stream? When you get a text? A picture? How do you even know what a realistic alien picture looks like, to know you've cracked the format? Is it in RGB, or maybe BRG and their sky is green like on Venus? How do you even know it's RGB and not CMY? Or what if that civilization sees in IR-Y-B like the common house cat? Your channels mistaken to be _their_ primary colours would yield a weird picture in which a lot of things (e.g., a rose or someone with a sunburn) appear to be literally red-hot. Did you even figure out right which fields are the width and height? Did you even get the header or chunk start in that millisecond slice of it, or are you working with some piece in the middle that relies on info you missed? Etc.
It's a monumental task to start from such a stream and even reconstruct the picture. Sure, you may relate better to visual stuff, but just getting that visual stuff back from the stream fragment is a monumental task that makes cracking the Enigma code look like a kindergarten exercise.
That's why I said send the prime numbers instead. Any civilization can count beeps. There is no advanced maths and data processing needed to decode that. Any average guy looking at that signal can go, "heey, the pulses come in groups of 2, 3, 5, 7, 11, 13... Jesus F. Christ, that's the prime numbers!"
You can send them the images later, after you get their answer. But the first step should be to make damn sure that it's a no-brainer to recognize the signal, what it is, and that it can't just be cosmic noise.
The guy has a point. A laser beam pointed straight up will sweep at _incredible_ speed over any receptor situated a couple of tens of lightyears from here. Even if that civilization were looking this way at the right time, had receptors strong enough for the task, had the luck of not having the beam blinded by our or their sun's light (there's a reason we have trouble detecting even Jupiter sized planets by their reflected light, which is higher than this laser will send), etc, it's something that will sweep over their sensor in milliseconds. At most you can say "oh, there's a bleep of light", but not even "oh, it's modulated". Much less have time to figure out what's being sent or how to decompress it.
And speaking of which, ffs, who got the stupid idea of sending encoded images? How about something as simple as morse codes, or train of pulses whose count are the prime numbers or Fibonacci's numbers? That's something that any civilization with even elementary maths knowledge and a primitive telescope can figure out quickly. "Hey, this can't be natural!" By comparison, a short faint burst of noise (which is what an alien data format would look like to you too) is likely to be written off as noise or as some unknown one-off cosmical phenomenon.
All in all it _is_ a stupid publicity stunt, and nothing more.
Just because something is described as "nanoparticle", it doesn't mean everything is created equal.
E.g., if it then disolves in water instead of staying a particle, the whole thing about it originally being a nano-particle doesn't mean jack squat by the time it's dissolved mollecules in your bloodstream. You could eat something containing, say, sugar nano-particles and it wouldn't do any more harm than the same quantity of sugar in any other form.
E.g., there are a ton of things already that do involve particle-sized bits of a substance in another substance it doesn't dissolve in. Mayo, for example. Emulsions are like that. The thing about mayo is that (A) we're talking _liquid_ particles, so nothing that would go poke a hole through your brain with its sharp edges, and (B) substances which are processed by your body long before that anyway. Making the droplets nanometre sized instead of micron sized wouldn't even start to make any difference. Sure, you might get a liver or cholesterol problem out of it, just like out of the same quantity of normal mayo, but none of it would be because of scary nano-particles.
Basically I wish both sides would just let go of the buzzword and get back to thinking critically.
On one hand you have the media and VCs, for which even shit sounds cool and high-tech if it even mentions nano-tech. Never mind that just grinding something into nanometre-sized dust particles is _not_ what nano-tech was supposed to mean.
And on the other hand you have the luddites for which the mere mention of nano-tech sounds like the Antichrist, and _must_ have the same effects as moon dust. Because, you know, a nanometre-sized drop of oil _must_ have the same effects on the body as a nanometre-sized piece of hard stone. It's all "nano" stuff, so it must all be the same, right? Well, wrong.
Basically let's just get back to thinking critically about it, and having some more actual information about it, instead of assuming that these guys' additives _must_ be either the second coming of Christ or the final sign of the Apocalypse, just because it has the magic word "nano" in there. Whatever it is, it'll get fed by the spoonful to lab mice, it'll get radioactively tagged, etc, and then they'll dissect those mice and see where that stuff ended and what happened to every single organ in those mice. If anyone finds a bunch of that in their brains, there'll be plenty of people screaming bloody murder, so you know to stay away from it. There'll be studies, counter-studies, astro-turfing, conspiracy theories and hentai fanfic about it. And then you can at least choose which you believe based on some more detailed claims than just its being "nano". There, it wasn't so hard, was it?
Disclaimer, I'm not even anti-Windows as such, but still:
1. The model capitalism was based on was #1 not #2. The ideal (19th century-style) capitalism idea is basically that of a market of commodities: all products are interchangeable, and a perfectly informed market decides which one offers the best bang/buck.
It's not just an ideological point of view. The role capitalism was supposed to serve was that of, well, basically the ultimate optimizer. Think, sort of, genetic algorithms. If the market needs X at all, there'll be tens or hundreds of competing entities trying to offer the best X at the best price possible. Each will make their own X1 variation at the best price they can manage, and the market will decide on the variant which fits the demand the best.
That was the strength of the western block over, say, the Soviet Union. That in the time the USSR planners decided always too late on what to produce and how much and in what way, the capitalist market could try a hundred ways and let the market choose the best one. That unless deciding "ok, we'll produce Volgas" which may be right or wrong, you can let a hundred people try a hundred different things, and end up with the Ford T1 model which was better and cheaper for most people.
Letting a corporation compete on how "toxic" they can be instead of competing on raw product merits is subverting that whole idea. It's, in fact, no better than the Soviet system. There too how "toxic" and subversive one could be (e.g., having high placed friends in the Party, or _being_ a party official who can send the opponents to Siberia) was most often what decided which model got produced and what got scrapped. We already know how well that went.
2. I wonder and worry about MS. (Or their managers, before someone accuses me of anthropomorphising... human managers.) Their whole history and practices shows that they're just not playing the same game we expect everyone else to play. And which, again, is the whole foundation of capitalism. Again and again, they seem more interested in just killing as many opponents as they can, as opposed to offering a better product or whatever.
Let me explain that better: it seems not even being toxic for survival reasons, but just being toxic for the hell of killing someone. Regardless of whether it's even a survival advantage or not. MS will even gladly take a huge loss (e.g., their XBox strategy) just to try to put someone else out of business. It's stuff that isn't even a survival advantage (making a loss never is), but just the sheer fun of killing someone just because they can.
Basically it's like watching a football game, where one of the players isn't even as much interested in playing the same game or even winning the game, as such, but just in kneecapping as many opponents as he can. Even winning (or losing) is merely a side-effect of killing or crippling everyone in the other team, rather than the goal and purpose of the exercise.
3. And I seriously worry about -- and am disgusted of -- the current US government's bending over to that kind of behaviour. Yes, that being toxic instead of competitive is an option for MS, is pretty obvious. But why tolerate such an entity? Not only it's condoning a major subversion of the very idea of capitalism, but... for _what_? MS is actually contributing very little to the US economy.
Microsoft is employing a grand total of 71,553 employees in 102 countries and regions as of July 2006. Total. Most of them actually support and sales/marketting and management people, and probably more than half off-shore anyway. Even at the scale of IT jobs, it's a spit in the bucket. Out of _millions_ of IT jobs, even after the exodus to India, we're talking maybe half a percent. At the scale of the economy, a helluva lot less.
So, while, yes, a government's priority should be keeping unemployment in check, MS is a spit in the bucket in that aspect. The effect of MS upon unemployment in the US is negligi
Slashdot Mirror
Except it rarely works that way. I have some experience with both chess and Go, and have been the inexperienced player in that scenario more than once. I've yet to see even one single instance where it works like that from beginning to end.
The inexperienced player may pull one or two surprisingly good maneuvers out of sheer dumb luck, maybe even gain a temporary advantage out of those. But in the long run he'll fail to use and consolidate that advantage and the more experienced player will _bury_ him.
The chance to win a match by sheer dumb clueless doing something random that the other isn't expected is negligible because it just needs too many moves in a row where that happens. If the chance to make a surprisingly good and unexpected move is, say, 1 in 1000 (remember, it has to be not just good, but also some radical new strategy that noone tried before and the good player isn't expecting), then the chance to make two in a row is 1 in 1,000,000. And the chance to make 4 in a row is 1 in 1,000,000,000,000. Keeping up like that for a whole game is just not going to happen.
Plus, good players are good because they can adapt and use logic to different situations. He's not going to just give up and run in circles for the next half an hour just because you did one different move. He'll keep reacting and probing and you only need to get out of that lucky streak once or twice for him to fully use it against you.
Basically "beginner's luck" is a myth. It's a crap excuse by people who aren't as good as they think, to not admit that they played badly. Or that maybe they let you win. But if they didn't, then that supposed beginner actually played pretty damn well.
And if Google's secret sauce is "beginner's luck", then maybe all it says is that the big "experienced" players are the ones playing badly in that space. Maybe it's not Google who's clueless there.
"They laughed at Einstein. They laughed at the Wright Brothers. But they also laughed at Bozo the Clown." - Carl Sagan
Did I mention fallacies? I think I did. In this case the Association Fallacy (sharing one quality or in this case one group of enemies with quantum physics doesn't actually prove this one true too), salted with a bit of Ad Hominem ("people like you...").
Here's a fun reality check for you: quantum theory actually had some actual experimental data to show, and made some easily falsifiable claims. (Falsifiable meaning it could have been easily proven false, if it were false.) And in the meantime it's shown a mountain of evidence and made possible practical stuff such as LASER or the Zener diode. _That_ is what makes it true, not sophistry like who laughed at it or who didn't. That's how science works.
Unfortunately that idea is lost on some CS clowns who think that just arguing personal beliefs, or occasionally preferences, is some form of science. And that expertise in some unrelated domain makes them automatically right. ("I'm the compiler guru, therefore I can tell you which methodology is more productive." Or "I'm the data structures guru, therefore I can tell you what's needed for intelligence.") No, it isn't. Show me the actual experimental proof, and then we'll talk. Until then, it's no better than the work of all those who wrote from positions of authority about what's needed to transmute lead into gold.
Believe it or not, _some_ things about it are known by now. There are more psychologists, neurologists, anthropologists and so on, all the way to stage "magicians", who've studied those various aspects, than CS clowns busy postulating they just know what's needed for intelligence. And the fun part is, the former group actually has some falsifiable data to show for their efforts.
At any rate, it doesn't matter if I'm arrogant (yes, I am) or what you choose to be incredulous about. Show me the experimental data and then we'll talk. That's how real science works.
Otherwise it ends up an "argument ouf ignorance" fallacy. When you've postulated that compression is _needed_ for intelligence, you've made a claim about the brain too. So it ends up basically, "We don't know how the brain on the whole works, so it must be true that it needed data compression to achieve intelligence." Which is a textbook example of that fallacy.
Your using a fallacy yet _again_? Dunno about "worse", but it looks pretty bad. What you assume about me still hasn't proved or disproved the theory you're busy defending. What _would_ even start to prove it would be at least _one_ example of an AI where pure data compression was essential to its workings.
Looking reasonable valid on the surface can be very mis-leading. Alchemy looked reasonably valid on the surface. Communism looked reasonably valid on the surface. Inteligent Design looks reasonably valid on the surface for a _lot_ of people. Etc.
Unfortunately, again, that's not how science works. Science works more along the lines of "show me the experiment". Show me that data compression is even useful at all in a working AI, before making broad claims that it's _needed_ for intelligence. It's that simple.
So, by the same logic, if you're not holding a Ph.D. in Economics, you're not allowed to say that an apology of Communism is bogus, right? They had plenty of Ph.D.'s in Moscow writing books about how great that system is. You wouldn't go and attack the work of people more educated than you out of sheer ignorance and stupidity, right?
Here's a fun concept for you: appeal to false authority. Because that's the fallacy you're committing there.
Show me someone who's actually built a working AI, and then we'll talk. That's a real authority I'm willing to accept any time. Anything else is, for the scope of discussing what's _needed_ for AI, just an unproven hypothesis. No more, no less. Even if their other maths is good, the supposition that's indeed needed for inteligent behaviour is no more than an unproven hypothesis pulled out of the ass. Arguing that since someone is an expert in the first, they must be automatically right about the second, is no more and no less than an appeal to false authority. It's like arguing that since Newton was good at physics, he _must_ have also be right about alchemy. (Which is another thing he studied, btw.) It's just that stupid.
So seeing that noone actually produced anything that can take a RL information stream and do anything even remotely intelligent with it, I'm not sure what makes them irrefutable authorities on the subject. Where _is_ the proof-of-concept program that uses data compression to actually extract useful knowledge out of Wikipedia and act intelligently about it? No, seriously. Anything else is just taking a guess. So exactly what there says that you should automatically stop thinking for yourself and take his unproven ideas for absolute truth?
In the meantime, however, we have this thing called a human brains that actually exhibits intelligence. And we're starting to understand some things that it does. Such as that, as I've mentioned, it does _not_ work by simply compressing the whole stream, but by extracting the relevant bits and utterly ignoring the rest.
We have _millenia_ of proof that if you're interested enough in the magician's left hand, you completely miss what his right is doing, or that if you focus on the miracle horse doing maths, you miss the owner giving it hints when to stop tapping. Heck, proofs that if they get your attention on something in the foreground, you'll completely miss the guy in a gorilla suit jumping up and down in the back, even made it onto mainstream TV.
So between some unproven hypotheses, and something that actually works that way, I don't know about you, but I'll go with what works here and now. I'll go with how the brain actually seems to work.
Here's some free, complimentary clue for you: "All that glitters is not gold." Just because someone wrote a ton of maths about compression, doesn't mean it's anywhere near applicable to building an AI.
And in fact, noone actually built yet anything even vaguely resembling intelligence yet, so maybe, just maybe, all those hypotheses are missing some vital point. Maybe, just maybe, just because you've compressed something to the minimum number of bits, doesn't mean jack squat about being able to use it effectively in real time. Maybe, just maybe, just because you've run arithmetic compression over a text, doesn't mean jack squat about actually being able to parse and navigate the semantic relationships in it. You just have a compression program and need some hours just to get your stream of bytes back, and be no wiser as to how to act intelligently on it.
In the meantime, however, we have a ton of psychologists, neurologists and, don't laugh, even stage magicians who've explored how the brain does work. You know, the only actual thing that actually exhibits some intelligence. We've worked out a lot of odds and ends about what happens there. We have some idea of what it filters out and when, how many seconds do some of the buffers hold the data, and so on.
So maybe you can take some hint from there. Maybe the information you need to extract and act on is really something like "[car][incoming]" instead of ivory-tower exercises like calculating the minimum number of bits you can compress the incoming stream to. That's the hard part: extracting the useful information there.
Or do you need something with lots of formulae? Read something about AI, lemming. There are a lot of people who've worked on various inference machines and such, and, surprise, none of them involved compression algorithms.
So, you know, speaking of "ubtracting out the sound of your own voice droning on at the drop of a pin about subjects you haven't made an effort to fully appreciate", maybe you can take your own advice there. Just because you get a boner about some maths, doesn't mean jack squat about applying it to a practical AI problem. Maybe start using your own head about how you'd actually act on that information instead. Just a thought, you know.
I can think of lots of stuff which would realistically count towards intelligence, but "compress knowledge" is the kind of thing that just sounds unbelievably stupid. In fact, it sounds like the kind of idiotic publicity stunt coming from a marketting guy who never heard of AI or programming.
The things you'd realistically need, and probably use every day unconsciously, include:
- good indexing: Just having the information somewhere in your head, well compressed is pointless if you can't quickly find it.
E.g., when you see a faucet, you need to immediately access such information as "turning it this way makes water come out, turning it the other way stops the water from coming out." You need to start from "faucet" and basically search for the information you need relating to "faucet". That's an easy example, but even trivial daily activities like driving your car often involve navigating a whole graph of associations, and if, say, a kid jumped in front of your car, you better not need hours to uncompress the data and find a solution.
- good "tokenization" so to speak. Think every word or concept being not a sequence of letters, but a pointer to that very concept. Think Wikipedia's hyperlinks. It may resemble compression on a superficial level, but that's like saying that a bird resembles the space shuttle on a superficial level. The goal here isn't just to minimize size by agnostically matching any chunk of text that came before, but to create a graph of interdependencies that you can navigate to find a solution.
E.g., again, if you started with "faucet" and went through "turn it clockwise", you need to imediately be able to reference "turn" and "clockwise" without sitting and sifting through a zip archive. And for that you need the real tokens, not any old chunk of matching text for compression reasons. A match for "turn" is good. A match for "rn i" between "turn it clockwise" and a previous occurence of "a barn in the backyard", as a compression algorithm would match, is utterly useless.
- good filtering. Yes, the brain is sorta lossy at all stages, because it mercilessly filters out everything that doesn't look relevant or in the focus of attention. E.g., when looking at the street, you might just see "it's a car", because the details about it don't look relevant at the moment. E.g., when focusing on the car for more details, you might completely miss the guy in a pink gorilla suit doing cartwheels in the background, because it looks irrelevant to your current focus of attention.
This is more important than it sounds, and is the _real_ compression for the most part when processing that data. The brain isn't just compressing the whole scene to save bandwidth, it's really mercilessly clipping 90% of it out, tokenizing the rest, and again clipping out most of the details about those tokens as long as they're not explicitly required. It's not just compressing the image of a car coming at you, it's transforming it into the "tokens" (so to speak) "[car] [incoming]". Anything else from that scene (e.g., the guy in a pink gorilla suit across the road) isn't compressed, it's filtered out, and so are the details about those tokens. (E.g., car model, colour, registration number, etc.) You can choose to pay attention to some of them (e.g., look at the registration number or the driver's face), and that will change the filtering criteria to give you those details at the expense of something else.
Even what you mention ("while my brain is good at remember the gist of knowlege, but really bad at losslessly recalling it.") is just an example of such filtering. A lot of the details were just filtered out before you even memorized that (e.g., you might have gotten the idea that a giraffe has a long neck, but not its colour, because it didn't seem important at the moment), an a lot just weren't stored or indexed or given a high priority. If it didn't look useful to remember what colour a giraffe is, it didn't get stored. Again, it's filtering, not dumb stream compression.
And so on.
And missing all that in favour of a "let's see who compresses the best" stunt seems just stupid and clueless.
Reminds me of when some other such database (which, as it turns out, powered our corporate firewall) flagged a big chunk of MS's website as drug-related IIRC.
If someone forced a collect call on me with no more warning than "someone is connecting your call", I _might_ call free() or delete() on their private parts.
And this, kids, is why you should use getters and (if needed) setters instead of making an Orkut-inspired design where everyone is "friends" with everyone they ever heard of. If anyone needs your private parts, they can ask nicely and you can give it to them in a civilized manner.
Sadly, this isn't slashdot-speciffic, but a general tendency of most people. People tend to not juggle too many variables about someone or something. ("John is a great coder, but he's an asshole, he's too introverted to manage a team well, and he's an average driver, and...") Even if they acknowledge such things as different factors, the subconscious tendency is to take the overall impression and apply it to everything. So if they like John a lot, they'll tend to view everything about him in that positive light. ("John is a great coder, and refreshingly frank and honest in dealing with people, even if that annoys some, has a refreshing hands-off approach to managing his team, and drives great, and...") If they dislike John a lot, they'll view everything about him in that negative light. ("John is an asshole who couldn't code his way out of a brown paper bag, is the most clueless PHB ever, and drives like a retard on hard drugs.")
People who look good and are well spoken tend to get the promotion, and as at least one study showed, they don't end up in prison as often. If you're in the jury and like that guy a lot, your natural tendency is to extrapolate that positive impression to everything about him. Such a nice guy _must_ be a honest and hard working guy, and probably got framed for that crime. If he's an ugly guy with a bad accent, well, the opposite applies. The motherfucker must be a no-good bum, and thank goodness we can put him behind bars.
The same applies to products. If they were made on the same chasis, with the same engine, and handled exactly the same, you'd probably still be inclined to say that a BMW handles better than Homer Simpson's dream car.
The same applies to companies. If you like a company a lot, then their products must also be the best of the best, their salesmen are honest, their employees are the most brilliant guys that ever walked the Earth, their patents are breakthroughs comparable to inventing the wheel, the waste they dump in the river is just pure water, their factories don't cause global warming, and their lawsuits are right and justified. That's why marketting and PR departments try to create a good image for the company as a whole.
And conversely, the same applies if you really don't like a company. Then their products by definition suck, their salesmen are backstabbing snake-oil salesmen, their employees are the rejects of vendor-machine-refilling school, etc.
It's not trolling, it's just extrapolating the general impression to each of the components. If someone would get a B grade on the whole, the tendency is to act as if every single component is the same B grade as the average.
It's not logical, but that's how humans work.
So some people apply that to MS. Whop-de-do. Humans acting like they're human on Slashdot. Who would have guessed?
"Adequately justified" is often just a matter of creating enough scary hype. See absynthe which is now legal again, after discovering that _no_ substance in it except the alcohol has an effect on the brain whatsoever. Yet a massive media campaign at one point presented it as a dangerous drug that turns ordinary people into crazed bloodthirsty serial killers and such bullshit. And enough hype and cerrypicking conjectures eventually were "adequate justification" for its prohibition.
Marijuana just had the mis-fortune of being hemp, and paper-from-hemp was at the time a threat to someone's paper-from-wood industry. (Yes, you've read it right. It had really nothing to do with people smoking it.) Unfortunately that someone also owned a media empire, and he used it to immediately unleash a barrage of fearmongering and bogus stories about crazed Mexicans going on murderous sprees after getting high on hemp. Even the name "Marijuana" started there, to make it further sound like a Mexican thing.
(Using the racism and nationalism card is a pretty common theme with most prohibitions and scare campaigns.)
At some point that was "adequate justification."
After that, and the ever increasing evidence that it doesn't actually _do_ any of that, it gradually seems to have backpedalled into:
- "yeah, but it's a gateways drug" (which not only was never proved, but also never answered the question "then why doesn't alcohol act as a gateway drug, if people are somehow made to progress to harder stuff?") and
- "yeah, but people end up stealing and mugging to support their drug addiction" (except it wouldn't happen if it was legal and reasonably cheap, like, say, tobacco is), and
- "yeah, but it's trafficked by mobsters and other criminals you wouldn't want in your neighbourhood" (who'd be out of job if it were legal), and
- "yeah, but once people start breaking the law to smoke pot, God knows what other crimes follow naturally." (The "gateway" argument all over again, and again missing the point that it wouldn't happen if it wasn't illegal to start with.)
More importantly it misses the point that Holland is an uncomfortable example of a country where it is still legal, and nothing bad happened. They're not overrun by gangs of drug-crazed psychos, and in fact (like most of continental western Europe) their criminality is actually quite low. It's not a gateway to harder drugs, either. Even assuming the "gateway" arguments wasn't bogus (it never was proved), if you can get your hemp legally, you're not that tempted to switch to something which can get you arrested instead.
But, eh, if you keep the hype campaign going, Jack Average won't think much further than what the idiot box tells him to think.
Think of it this way: look at how many bits the human DNA has. Each nucleotid pair is one of 4 possible values, so 2 bits. A human has about 3 billion pairs. So on the whole about 750 megabytes or so.
I dunno about you, but I'm thoroughly impressed. _You_ try programming a human in 750 MB, and then you can criticize. I'm talking not only the brain (which is a feat by itself), but also the whole organism there, including immune system, self-healing, metabolism, etc.
I don't even know if it's unstructured. What we have here is some people trying to use that stuff without knowing what it really does or how it was supposed to be used. It's sorta like watching the client PHB trying to just drag and drop a button on a form himself, and then wondering why it doesn't work like he just assumed it would. ("Hey, I dropped a "Save" button there and it doesn't actually save when I click it!") You can't really blame it on the original programmer if it doesn't work that way.
So if I were mean, I'd up the ante and also say "and program it so it also works when some clueless guy later tries to use a human cell in a mouse, in a way that was never in the specs." But let's be generous and skip that. Just program a human in 750 MB, no matter how.
What _our_ engineers (and I'm one, so I'm allowed to criticize) manage today in 750 MB is a stupid text editor or a spreadsheet. We're past even just structured programming. Now we pack everything in layers upon layers of frameworks, EJBs, factories, decorators, managers (the pattern, not the PHB), events, SOAP, JMS, etc, just because it's _fashionable_ to have one more buzzword on the resume. And, oh, if it's Java, let's add add a layer of introspection too, because it's become soooo unfashionable to just write "myObject.getID()" instead of cracking the class open the class at runtime and reaching in its internals.
Let me stress again: this crap doesn't even have to do with OOP or with structured programming any more, and in fact sometimes _prevents_ proper OOP. Project after project I run into crap designs where you _can't_ use OOP, e.g. define a simple subclass of something, because for example there's an EJB layer in the way and the other end wouldn't know how to deserialize your modified objects. Or because all the data objects are generated from some XML definitions -- and I don't just mean the stuff that'll get persisted in the database, but really internal data objects -- just because someone thought it's _cooler_ to write an XML and run a third-party generator than to just write the fucking member definitions and ask Eclipse to make getters and setters for them. And as a result you can't even attach the relevant methods to them, like you learned in the OOP classes, like attaching a "findChild()" method where it belongs in the tree node, because it would get overwritten in the next build when those objects are generated from XML.
Engineering used to be about having a problem and designing the best solution to it. E.g., you have a river to cross, and you consider whether it's best to build a bridge, or a tunnel, or a ferry, and pick the simplest and cheapest solution that solves the problem. Now we're at the stage where we want to have a "suspension bridge" buzzword on the resume, and we'll cheerfully dig a canal just to have something to build that bridge over. Or detour the road through 3 states to the side so we can find a gorge to build the bridge over. Must have that precious buzzword even if it kills the project. That's not engineering, that's marketting and playing.
Anyway, at this rate we'll soon need a 750 MB framework just to pass the parameters around.
Except it's not even actually funny. There are funny reviews, there are good no-holds-barred "look at all that's bad with these games" reviews (e.g., on Something Awful), and both are good and refreshing in a world of bending over to EA's marketting money. But this is neither. This is one guy ranting incoherently, while trying to sound like a hip wisecracking smartass, but failing miserably.
It's just trying to construct some convoluted mixed-metaphor insults that go on for a paragraph. And more often than not are of the kind of "if you dare disaggree with me, then here's in how many ways you're a retard huffing spray can and probably also have fun banging your had against the inside of a discarded refrigerator." I've read
Basically what I'm saying is, insult the game if you must, but if you can't make a point without ad-hominems then you probably don't have a point to start with.
They say that if you try too hard to make an impression, that's the impression you'll make. And this is proof of it, if anything. The guy is trying too hard. I'd rather have had him do a professional review of why a game is bad, and keep the wisecracks short and to the point. A one-liner to make the point is good. A two-paragraph convoluted argument about why the reader is gay and a retard to boot, out of a maybe 8 paragraph "review", is just a piss-poor substitute for actual review content.
And whatever content it had, it's mostly picking on some 10 to 15 year old games and whining that they don't have the graphics/content/controls of a 2003 game.
Some of the screenshots there are from games from the 80's and, frankly, don't look any worse than any other game released in the 80's. Sure, they look bad by today's 3D pixel-shaded bump-mapped ragdoll-animated standards, but back then pixelated midgets and 4-colour backgrounds were pretty much the norm. Ditto for the animations: you had maybe 1 or 2 attacks for yourself, and maybe 1 or 2 for the enemies, and that was that. Cloned enemies repeatedly kicking or throwing purses at you were the norm. You just didn't have the memory to store detailed life-like backgrounds and a gazillion life-like animations.
And yes, occasionally punches and kicks went right through a character's chest, because you didn't have 3d ragdolls to stop a kick exactly where it touches the enemy. If your sprite of a character had his leg kicking at a full 3 ft distance and the enemy was 2 ft away, yeah, your foot would go right through them.
Some seem to be from the age when everyone discovered they have just enough RAM to use scanned photos for the characters. In many ways it was a worse age than the aforementioned pixelated-midgets age. At any rate, it still had the problems of the previous age, with an uncanny valley factor thrown in. Now you had realistically scanned enemies acting like stiff cardboard cutouts. Or in some cases you just had blurry contrast-less scanned midgets instead of the previous pixelated ones with good contrast.
Don't get me wrong, I'm not saying they were the great games of their era or anything. But it seems kind of silly to pick on the graphics of an 80's game just because they're 80's kind of graphics.
Very true, but I do know someone who was left for a whole year to drive a company into the ground before he was fired. Not just non-work, but he caused all programmers and designers to quit together, he alienated all the business partners that company ever had, etc. (He was a complete asshole to employees, customers _and_ the companies he subcontracted to.) You'd think it would be obvious that something is wrong by the time he has no more employees he can make the programs with, and has to subcontract everything, but it obviously wasn't obvious to his superiors.
So, again, just because something or someone is counter-productive, it doesn't mean it won't happen. Sad, but true.
Not every data entry typist has access to the credit card numbers. (And if all do, then there's something wrong with your database and applications.) There are jobs as boring as just correcting the format of entered telephone numbers, for example.
This is actually the mis-conception that bothers me the most, and sadly it's been hammered into a lot of otherwise intelligent people's heads by decades of propaganda and lies. Everyone thinks that if someone does a joint or two a day, they're inherently becoming slackers, thieves, maybe murderers, and they'll probably switch to harder drugs soon too. And that's just not true.
I don't do drugs myself (I figure I'm already ruining my lungs enough with tobacco anyway, plus I like to stay Lawful Good), but I've worked with people who smoked pot. They were no different from any other human. In fact, I've worked for a couple of years with two of them before finding out they did pot, and I didn't suspect a thing.
No, they weren't unreliable. No, they didn't sell company data. (And the two I've mentioned above were the DBAs, so they had access to literally everything in our databases, if they wanted to.) No, they didn't steal anything. No, it wasn't a gateway to hard drugs.
If anything, in retrospect they were calmer and more professional than some other coleagues, although I don't know if that's due to pot or that was just their personality type. They just did their job in situations where I'd have been at someone's throat for doing something stupid. But again, it might just be their normal personality for all I know.
And frankly it bothers me to see such people being villified and victimized. It bothers even more to know that a chunk of my tax dollars goes into victimizing such people.
Folks, if you want to argue physics, go read a physics book first. Wishful thinking and appeal-to-hidden-motives phalacies are _not_ how science works. Show me some real physics data, not rants about GM and Exxon conspiracies, not copious hand-waving, and not verbal fallacies. Those do not quite a scientific proof make. I don't freakin' care what GM or Exxon think. Try reading at least a high-school physics book and using your own head for a change.
Again, let me enlighten you: an electric motor is made of coils through which current passes. Big thick coils consisting of layers upon layers upon layers of wire, like big spools of thread, on a ferrite core. That creates the magnetic field that creates the torque. (Actually, lemme play it safe and translate "torque" for you too: it's what makes it turn.)
_That_ is the wire I'm talking about. That mile of wire in those coils. Not the battery cables.
So, yes, the battery cable doesn't mean shit. Too bad your knowledge of electric motors means just about as much too.
Ah, way to start a good old fallacy. You can't actually make your point without attacking the other's supposed motives, can you? Well, anyway, let's ignore that and have a look, shall we?
"Ignoring friction" is good and fine in an ideal world, but in the RL it isn't. Heck, why stop at friction? Let's ignore all resistance for that matter, and then you don't even need an engine to keep going.
Heh. Nope. Funny, but just nope. An electric motor provides pretty much constant torque at any RPM, and regardless of how that RPM varies. Pretty much the torque (hence acceleration) limit for a given motor is how much current you can run through it. If anything, running more current through it just means more losses, so you're really saving nothing by flooring the "gas pedal" instead of starting slowly. Au contraire.
Ah, again ignoring RL. _If_ you had a perfect system, where nothing is lost, the world would be such a wonderful place indeed. Unfortunately RL doesn't quite work that way.
Plus, you'd save exactly as much there with a hybrid car using an ultra-capacitor, so it's hardly anything specific to going all electric.
Heh. How about giving me some numbers, then? That big ol' combustion engine _and_ the fuel tank still weighs a fraction of what the batteries would weigh. Let me do some maths for you:
Gasoline stores some 14kWH/kg. About a quarter of it is actually useful, and about 75% are losses of the that big ol' combustion engine. So if you filled your tank with, say, 20 kilos worth of gasoline (about 25 litres, or about 8 gallons), you'd have about 280 kWH stored there, out of which some 70 actually help move your car.
For the best Li ion batteries available nowadays, the density is 120 WH/kg, or 0.120 kWH/kg. (For lead-acid it's 0.04 kWH/kg. For ultra-capacitors it's 0.003 to 0.005.)
Let's also assume that you have an ideal system where the electrical car is 100% efficient. It isn't, but let's pretend for a bit, shall we? So for this ideal car we'd need to store only 70 kWH to be equivalent to the above quantity of gasoline. Right?
So you'd need 70 / 0.12 = 583 kilos worth of batteries. I don't know what car you drive, but most internal combustion engines are lighter than that.
To put things into perspective, for an Audi A8 4.2 TDI Quattro the engine weighs 255 kg. And here we're talking a big freakin' 4.2 litre 8-cylinder dual-turbo Diesel engine. The 4.2 litre 8-cylinder gasoline version weighs in at 190 kg. The BMW M54 engine, a big ol' 3.0 litre 6-cylinder engine, weighs 170 kg. A Ferrarri F50's engine weighs 198 kg, and that's a big ol' V12. But these are powerful sports engines. Your average small commuter car will have a much smaller engine.
But so far we've been making the comparison based on the flawed assumption that the electric engine weighs n
Which is still rather irrelevant if you use a coal motor (ok, steam turbine) to produce that electricity in the first place. No matter how efficient you make that electric motor, the efficiency of it all is still basically the product of that _and_ the turbine that produced it in the first place.
Which, I do believe, was the GP's point.
Like... nuclear power plants, for instance? Because everything else either pollutes (e.g., coal plants) or just doesn't quite scale like some people seem to think. If you think you could just run whole countries _and_ their cars off wind turbines, solar energy and hydro plants, I want to know what planet you live on.
I.e., again, you actually just make the GP's point. Either you charge that car from a nuclear plant, or you've just moved pollution somewhere else. That was the whole point. Being both anti-nuclear and anti-pollution is just an unattainable proposition, short of returning to the stone age.
Unfortunately, a large part of the reason there is nuclear energy, so again you just make the GP's point.
Another reason is that electricity consumption didn't rise that much either. Most of the manufacturing industry actually migrated _out_ of the USA, so industrial consumption actually dropped, helping somewhat offset the rises in other parts. And home consumption didn't rise that much either. And to stay "on topic" you don't yet have every Jack Redneck driving an electric pickup-truck and Jane Soccermom driving an electric SUV, either.
Heh. If anyone knew how to make a filter that stops CO2, we wouldn't be having the whole global-warming hysteria in the first place. So _how_ are you going to manage that? No, seriously, I want to know.
Rest assured that burning coal at a power plant to produce power for an electric car releases would release _exactly_ the same quantity of CO2 into the atmosphere as burning it directly in your car... if distribution, batteries and electric motors had exactly 0% losses. As it is, it actually releases _more_. Ah, but a car runs on gasoline, not on coal, so it's already getting better for the cars.
Ditto for dispersion. How are you going to keep the CO2 from dispersing into the whole atmosphere? No, really.
So, again, you just make the GP's point. Either you charge that car from a nuclear plant, or no, we don't actually know how to manage CO2 pollution from factories either.
Which is irrelevant to the problem being discussed, so I won't get into that either.
Yes, it _can_ accelerate faster, but that's irrelevant to the pollution topic. Well, not quite: accelerate faster == use more power.
Yes, it can even get some of its energy back when braking, but that's losing sight of the fact that the electric car has to carry half a ton of batteries too. So it will actually expend extra power into moving all those batteries (doubly so if you want to move them faster), which means extra pollution at some power plant down the line. Unless, of course, you use nuclear power.
Unfortunately that just boils down to "it could get better in the future" optimism. Yes, battery chemistry will probably eventually improve, but here and now it still sucks big fat hairy arse compared to gasoline. Here and now driving an electric c
Actually, that's not entirely correct. No star we know produces elements heavier than iron and nickel, which aren't very radioactive. In fact, they're the most stable nuclei we know.
The thing is, anything lower than iron and nickel tends to release energy when fused into something heavier. Anything heavier than that needs to absorb energy to fuse into something even heavier, and conversely releases some energy when split.
So eventually the reaction stops at iron and nickel. Given intense photon bombardment in the star, most nickel actually disintegrates right back into smaller nuclei, not fuse further into heavier stuff. Iron pretty doesn't do anything whatsoever, and just stays iron.
The thing there is that as you move upwards, the energy and temperature requirements tend to be insane. For example for the next step up from fusing hydrogen into helium, it takes a red giant and temperatures of about 100 _million_ Kelvin to even fuse helium into carbon before blowing itself up.
And most stars either (A) stop short of even that and become a red dwarf, or (B) blow themselves up within seconds when they start fusing helium, because that's a very unstable reaction, whose rate increases with temperature, and temperature increases with fusion rate.
But at any rate, even if you had a star massive enough, you wouldn't get many nuclei past iron, or you wouldn't get them out of the star. By the moment a star got massive and hot enough to start fusing iron into something heavier, it would just rapidly lose heat in that reaction. It just can't explode that way, so at most you'd get a black hole in the end of it all.
So since you mention stars exploding... well, that's actually where the heavier elements come from. Supernovae don't just spread those heavier metals, they _create_ them. The iron, carbon, helium and whatever else was created will be smashed with tremendous amounts of energy and at insane temperatures, and a lot of it will fuse into heavier stuff. And since the star is already blowing up, they'll get spread all over the place.
1. Some forms of monitoring are actually dirt cheap.
To start with the obvious, spyware is pretty ubiquitous at some companies, and that includes company laptops. So then people take them home and use them too for IM, slashdot, VOIP, updating their "anonymous" blog, and whatnot, and you can see where that is going.
E.g., someone posted a while ago, in a thread about tele-commuting, about how he knew an employee wasn't really working at home because he looked on XBox live all the time and after a couple of weeks the employee had 5 achievements in Oblivion. (Never mind that Oblivion is a game which can be finished in a weekend if you just follow the main story, or in a week without telecommuting even if you do every single side-quest. And 5 achievements aren't really that much.) That's a form of surveillance.
Google can also be used as a cheap form of surveillance, because most people don't really try to be anonymous. Or can be identified by details they provide.
Cell phones can also be tracked, as proven by a recent article, but I didn't bookmark it. Basically a journalist used such a tracking service on his girlfriend's phone. It asked for confirmation once at the start, and from there it was basically in stealth mode. In that case it was with her knowledge, for research purposes, but you can see how that can happen without knowledge too, if you have access to a "logged-in" phone for a couple of minutes. Company cell phones are a prime example: they can be subscribed to tracking before you even get the damn thing.
2. The line of reasoning that something won't happen because it's not making any money (or preventing losses) for the company is flawed too, and assuming that humans on the whole only do perfectly rational stuff supported by solid logic and numbers. That's false. Humans do a lot more for emotional reasons than for anything even vaguely resembling cold logic supported by facts.
Some PHBs (A) have nothing better to do with their time (even doing lunch and painting powerpoint foils only takes so much time), and (B) are complete control freaks. They don't do it because it actually helps the company in any form or shape, but just to feel in control of something they actually don't really know how to manage.
Even HP's case, if you look at it, is really no more than some control-freak exercise. If you look at the "leaks" they were investigating, the grand acts of treason to the press so to speak, the mind boggles. One executive had unauthorizedly told the press that he's tired after a long board meeting. Or that HP hopes to sell more of their Opteron servers in the future. (Well, of course. Is their any company who actually hopes to sell less and lose market share?) It's benign, uninformative and bloody useless small talk, not any actual company secrets.
But someone was chuffed that a director dared talk to the press at all, even such uninformative small-talk, without their royal seal of approval. I.e., a control freak. That's really how that espionage and stalking affair got started.
3. Even when logic and facts are involved, a lot more often than not, the goals are PR, looking good, etc, not "is it making the company money." You can see it from company policies and politics to PHB's more concerned with maintaining an illusion to their superiors than with managing what they're supposed to manage. Whole man-years get spent on just seeming to do something about a problem, instead of just fixing it.
Or to take your example with drug testing, the thing is: people aren't testing only investors and board members. You know, people who could actua
See, before humans figured out how to smelt iron out of ore, there were weapons made out of meteorite iron. A certain number of meteorites are nearly pure iron, and better yet, some is even already alloyed with stronger metals. They were rare and more expensive than gold, but it was a weapon which could pierce right through a bronze cuirass, and was often credited with magical properties. Kings and nobles paid a small fortune for them.
Some of the myths around that kind of equipment persisted even after it was known how to just smelt iron ore. E.g., the celtic myths about cold iron against elves. The only iron that can be processed without heating from start to finish is, you guessed, a chunk of stuff that was weapon-grade iron from the start, not ore. That's more often than not a meteorite.
So other than maybe modern times and construction crews with bulldozers, you wouldn't just throw away such a rock if you found one. You'd sell it to a smith for a small fortune, and he'd make a weapon for a king and sell it for a bigger fortune.
Dunno about Apple, but if I were mass producing those things, I would _not_ build the thing empty, connect it to a Mac by hand, transfer the stuff to it slowly via Firewire, etc. That kind of "let's connect a cable, launch this handy application and click here to transfer the files" is ok for a mom-and-pop shop, but when you're mass producing stuff you just want to shave the last penny off the manufacturing costs.
So the way it's done is you take the working prototype, make an image of its hard drive, and write that on every hard drive before it's even assembled into the iPods.
Think, basically, how your IT department doesn't come with a suicase full of install CDs for Windows, Word, etc, for each PC. They just make an image off one workstation and then install that on all others. Much faster.
Same thing here, only more automated.
So if that image was made from a HDD with the virus on it, the assembly line will mindlessly churn thousands of copies of that.
That's actually the problem. Your average PHB indeed doesn't know jack shit about the difference between ReiserFS or FAT, or between Java and Visual Basic. So he'll take that kind of decisions not based on their actual merits, but based on rumours, over-simplified half-truths they half-understood from some IT-for-managers ragazines, fashion, and what the nice MS/IBM/whatever salesman filled their head with during a round of golf.
I've seen people actually take such stupid decisions as "let's use a single-user database and just copy the database file on the department's file server", in that case MS Visual Fox Pro for a reason as stupid as "Visual Fox Pro is more visual than Java". Once the nice MS salesman showed them some dragging and dropping buttons around (and, as everyone knows, there's nothing else to programming an app than dragging and dropping the buttons on forms), any other considerations like concurrent access, transactions, available tools and libraries, etc, went right over their head.
So the danger is precisely that at some point a nice salesman shop drops by and goes "whoa, you guys run SuSE? Did you know they paid a convicted murderer to develop their filesystem? Every time you save your powerpoint presentations on that file server, you have an innocent's blood on your hands, not to mention all over your neatly formatted presentation. Now if you upgraded to Vista Super-Professional Snake-Oil Edition, you'd show your support for the Bill and Melinda Gates Charity and be _much_ more fashionable among your peers."
No. Just no. See, the world is full of idiots doing something stupid and pointless. Those are a dime a dozen, and comparing them to people like Columbus or Magellan is just insulting to the latter category. The ones who changed history, e.g., Columbus or Magellan, weren't retards doing publicity stunts, they were people who put some solid thought into what they were trying to do.
E.g., Columbus's calculations might have been wrong, but he started from solid evidence that the Earth must be round. The idea that travelling West can get you somewhere in the East was sound. If there hadn't been this unknown continent in the way, he would have actually arrived in Asia as he planned.
E.g., Magellan actually did even better. He set a goal, put some solid thought into it, and actually achieved it.
Do you understand? That's the difference between those and Yahoo's publicity stunt. In the comparison to Columbus and Magellan, Yahoo comes out just about comparable to the village idiot running around with pencils up its nose. Sure, it gets some attention and maybe even some money, but comparable to Columbus and Magellan it ain't.
That is, assuming Hanlon's Razor applies. ("Never attribute to malice that which is adequately explained by stupidity.") But there's also a chance that it's just plain old dishonest, and they don't actually give a shit about actually achieving anything except publicity. I.e., pretty much as ic Columbus had just stirred up a frenzy, got some money out of it, then went and got drunk in France with that money. Without even giving much of a damn about actually reaching the Indies. Hey, he did his publicity stunt, got his money, the rest isn't his problem.
If you gave an alien a photo, yes. But if you give them a signal encoded with some proprietary lossy compression? Heh.
The problem isn't the photo. The problem is that you get a stream of 1 and 0 and you have to figure how the fuck to even make heads or tails out of it. Before you'd get a photo to use your alien intelligence on, you'd have to start from scratch in, basically, cracking the encryption. Because that's what a JPEG is to someone who never heard of the format specs: as good as encrypted.
And you don't even have any known cleartext to aid you in breaking that code. What are those 1s and 0s trying to tell you? When you should stop trying new things on that stream? When you get a text? A picture? How do you even know what a realistic alien picture looks like, to know you've cracked the format? Is it in RGB, or maybe BRG and their sky is green like on Venus? How do you even know it's RGB and not CMY? Or what if that civilization sees in IR-Y-B like the common house cat? Your channels mistaken to be _their_ primary colours would yield a weird picture in which a lot of things (e.g., a rose or someone with a sunburn) appear to be literally red-hot. Did you even figure out right which fields are the width and height? Did you even get the header or chunk start in that millisecond slice of it, or are you working with some piece in the middle that relies on info you missed? Etc.
It's a monumental task to start from such a stream and even reconstruct the picture. Sure, you may relate better to visual stuff, but just getting that visual stuff back from the stream fragment is a monumental task that makes cracking the Enigma code look like a kindergarten exercise.
That's why I said send the prime numbers instead. Any civilization can count beeps. There is no advanced maths and data processing needed to decode that. Any average guy looking at that signal can go, "heey, the pulses come in groups of 2, 3, 5, 7, 11, 13... Jesus F. Christ, that's the prime numbers!"
You can send them the images later, after you get their answer. But the first step should be to make damn sure that it's a no-brainer to recognize the signal, what it is, and that it can't just be cosmic noise.
The guy has a point. A laser beam pointed straight up will sweep at _incredible_ speed over any receptor situated a couple of tens of lightyears from here. Even if that civilization were looking this way at the right time, had receptors strong enough for the task, had the luck of not having the beam blinded by our or their sun's light (there's a reason we have trouble detecting even Jupiter sized planets by their reflected light, which is higher than this laser will send), etc, it's something that will sweep over their sensor in milliseconds. At most you can say "oh, there's a bleep of light", but not even "oh, it's modulated". Much less have time to figure out what's being sent or how to decompress it.
And speaking of which, ffs, who got the stupid idea of sending encoded images? How about something as simple as morse codes, or train of pulses whose count are the prime numbers or Fibonacci's numbers? That's something that any civilization with even elementary maths knowledge and a primitive telescope can figure out quickly. "Hey, this can't be natural!" By comparison, a short faint burst of noise (which is what an alien data format would look like to you too) is likely to be written off as noise or as some unknown one-off cosmical phenomenon.
All in all it _is_ a stupid publicity stunt, and nothing more.
Just because something is described as "nanoparticle", it doesn't mean everything is created equal.
E.g., if it then disolves in water instead of staying a particle, the whole thing about it originally being a nano-particle doesn't mean jack squat by the time it's dissolved mollecules in your bloodstream. You could eat something containing, say, sugar nano-particles and it wouldn't do any more harm than the same quantity of sugar in any other form.
E.g., there are a ton of things already that do involve particle-sized bits of a substance in another substance it doesn't dissolve in. Mayo, for example. Emulsions are like that. The thing about mayo is that (A) we're talking _liquid_ particles, so nothing that would go poke a hole through your brain with its sharp edges, and (B) substances which are processed by your body long before that anyway. Making the droplets nanometre sized instead of micron sized wouldn't even start to make any difference. Sure, you might get a liver or cholesterol problem out of it, just like out of the same quantity of normal mayo, but none of it would be because of scary nano-particles.
Basically I wish both sides would just let go of the buzzword and get back to thinking critically.
On one hand you have the media and VCs, for which even shit sounds cool and high-tech if it even mentions nano-tech. Never mind that just grinding something into nanometre-sized dust particles is _not_ what nano-tech was supposed to mean.
And on the other hand you have the luddites for which the mere mention of nano-tech sounds like the Antichrist, and _must_ have the same effects as moon dust. Because, you know, a nanometre-sized drop of oil _must_ have the same effects on the body as a nanometre-sized piece of hard stone. It's all "nano" stuff, so it must all be the same, right? Well, wrong.
Basically let's just get back to thinking critically about it, and having some more actual information about it, instead of assuming that these guys' additives _must_ be either the second coming of Christ or the final sign of the Apocalypse, just because it has the magic word "nano" in there. Whatever it is, it'll get fed by the spoonful to lab mice, it'll get radioactively tagged, etc, and then they'll dissect those mice and see where that stuff ended and what happened to every single organ in those mice. If anyone finds a bunch of that in their brains, there'll be plenty of people screaming bloody murder, so you know to stay away from it. There'll be studies, counter-studies, astro-turfing, conspiracy theories and hentai fanfic about it. And then you can at least choose which you believe based on some more detailed claims than just its being "nano". There, it wasn't so hard, was it?
Disclaimer, I'm not even anti-Windows as such, but still:
1. The model capitalism was based on was #1 not #2. The ideal (19th century-style) capitalism idea is basically that of a market of commodities: all products are interchangeable, and a perfectly informed market decides which one offers the best bang/buck.
It's not just an ideological point of view. The role capitalism was supposed to serve was that of, well, basically the ultimate optimizer. Think, sort of, genetic algorithms. If the market needs X at all, there'll be tens or hundreds of competing entities trying to offer the best X at the best price possible. Each will make their own X1 variation at the best price they can manage, and the market will decide on the variant which fits the demand the best.
That was the strength of the western block over, say, the Soviet Union. That in the time the USSR planners decided always too late on what to produce and how much and in what way, the capitalist market could try a hundred ways and let the market choose the best one. That unless deciding "ok, we'll produce Volgas" which may be right or wrong, you can let a hundred people try a hundred different things, and end up with the Ford T1 model which was better and cheaper for most people.
Letting a corporation compete on how "toxic" they can be instead of competing on raw product merits is subverting that whole idea. It's, in fact, no better than the Soviet system. There too how "toxic" and subversive one could be (e.g., having high placed friends in the Party, or _being_ a party official who can send the opponents to Siberia) was most often what decided which model got produced and what got scrapped. We already know how well that went.
2. I wonder and worry about MS. (Or their managers, before someone accuses me of anthropomorphising... human managers.) Their whole history and practices shows that they're just not playing the same game we expect everyone else to play. And which, again, is the whole foundation of capitalism. Again and again, they seem more interested in just killing as many opponents as they can, as opposed to offering a better product or whatever.
Let me explain that better: it seems not even being toxic for survival reasons, but just being toxic for the hell of killing someone. Regardless of whether it's even a survival advantage or not. MS will even gladly take a huge loss (e.g., their XBox strategy) just to try to put someone else out of business. It's stuff that isn't even a survival advantage (making a loss never is), but just the sheer fun of killing someone just because they can.
Basically it's like watching a football game, where one of the players isn't even as much interested in playing the same game or even winning the game, as such, but just in kneecapping as many opponents as he can. Even winning (or losing) is merely a side-effect of killing or crippling everyone in the other team, rather than the goal and purpose of the exercise.
3. And I seriously worry about -- and am disgusted of -- the current US government's bending over to that kind of behaviour. Yes, that being toxic instead of competitive is an option for MS, is pretty obvious. But why tolerate such an entity? Not only it's condoning a major subversion of the very idea of capitalism, but... for _what_? MS is actually contributing very little to the US economy.
Microsoft is employing a grand total of 71,553 employees in 102 countries and regions as of July 2006. Total. Most of them actually support and sales/marketting and management people, and probably more than half off-shore anyway. Even at the scale of IT jobs, it's a spit in the bucket. Out of _millions_ of IT jobs, even after the exodus to India, we're talking maybe half a percent. At the scale of the economy, a helluva lot less.
So, while, yes, a government's priority should be keeping unemployment in check, MS is a spit in the bucket in that aspect. The effect of MS upon unemployment in the US is negligi