I was monkeying around with a C64 emulator the other day, and it struck me how bad those old OSes were. I do have some nostalgia for these things, but more for the times they represented in my life than because I miss the hardware and software. In truth they were mostly cobbled-together messes.
BeOS is the only one I truly miss, and that is because it had something none of the current OSes have: Low user latency. With the current crop of OSes we take it for granted that:
machines take minutes to boot,
applications take tens of seconds to launch and quit,
opening or changing a filesystem view takes a second or longer,
applications can completely freeze out the user for tens of seconds at a time (like when a browser can't connect, or when Mac OS X can't find a disk volume it thinks should be there),
all user actions are accompanied by delays, even ones that would be trivial to avoid with prefetching or caching
What I miss about BeOS was the whole design aesthetic of putting the user first, never blocking user input, and making the common use cases fast.
As an experiment in grad school, for about 6 months I completely switched over to Dvorak. It wasn't about becoming a faster typist; I'm a pretty capable typist on QWERTY (> 100 wpm), and in a practical sense I find that thinking of what to type is usually the rate-limiting factor. I had two goals: (a) See whether and how quickly my brain could retrain, and (b) See whether there was any impact on some mild pain in my wrists (I was doing a lot of typing).
As others have pointed out, with keyboard remapping it's easy to get Dvorak on any computer you might happen to use. So in a practical sense there is no longer any lock-in of that type.
My experience:
After a few weeks, I was nearly as fast with Dvorak (probably around 70 wpm) as I'd been with QWERTY, and making continuing gradual progress.
There is noticeably less hand movement with Dvorak than with QWERTY, when typing standard English text. Whether this translates into higher ultimate speed, or less repetitive stress injury, is unclear to me.
The better I got with Dvorak, the worse I got with QWERTY. After a while it felt very odd to go back to a QWERTY keyboard. I think you'd really need to work at it to be simultaneously competent with both layouts.
Dvorak is pretty suboptimal with respect to common keyboard shortcuts. Things like ctrl-x to cut and ctrl-v to paste have been put in locations to make them comfortable on QWERTY, but Dvorak generally remaps shortcuts, control keys, etc. into inconvenient locations.
Both are pretty suboptimal for a lot of the typing I actually do, like coding. My ideal keyboard layout would have (){} on home row.
Ultimately I switched back to QWERTY for the keyboard shortcut issue noted above. Dvorak didn't have nearly as much impact on my wrist pain as taking breaks from work did.
My conclusion is that QWERTY vs Dvorak is largely irrelevant. There may have been more relevance before computers, when a lot of typing was done as dictation or copying. However most people today type as they're composing the words, and I strongly suspect brain speed is the practical limiting factor, not keyboard layout.
Well, the short answer is that the world needs more than one financial instrument for the same reason it needs more than one programming language, operating system, microprocessor variety, animal species, type of food, and so on. It's because diversity is inherently good. (Except when it isn't, which is pretty rare. Sometimes the benefit of standardization trumps diversity, e.g., electrical plugs and rail gauges.) Diversity creates competition, and opens up room for innovation. It also makes a system more, not less, resilient to catastrophe.
If you're looking for something to blame for your vanishing 401(k), here is my starter list:
Overly-lax regulatory environment, specifically poor reporting requirements for financial firms, resulting in little visibility into their real risk exposures.
Rampant moral hazards throughout the financial world, most notably in mortgage origination. (I.e., your mortgage broker and underwriter don't care a dime if you default or not. Odd, considering they're making the decision of whether to give you the mortgage.)
The Fed's stubborn refusal for years to recognize asset inflation as a valid form of inflation, instead using CPI as the exclusive metric for determining monetary policy.
Simplistic financial modeling by the quants on Wall Street, which treated mortgage defaults as uncorrelated events from one consumer to the next. (I believe this is what TFA was trying to get at, albeit poorly. In defense of the quants, it's probably impossible to predict when the complex nonlinear system that is our economy will tip into a strongly correlated regime as we've seen recently. Their dilemma is how to incorporate a known, but unquantifiable, term into their quantified decision models.)
Fear of recession among our elected officials, which induces them to be over-aggressive in counteracting natural downturns (small recessions), therefore allowing stress to build up in the system. They'd rather have a 5x recession on the next guy's watch than a 1x recession on theirs. Recessions are like avalanches at a ski resort: They are unavoidable, and the goal should be more frequent, smaller corrections.
A good old-fashioned crisis of confidence among the people. Honestly I'm spending less than I used to, in case I lose my job or get a pay cut. The problem is everyone is thinking the same way, which is guaranteed to depress aggregate revenue (GDP). It becomes self-fulfilling.
I doubt that financial product complexity is on the list. I'm also skeptical that greed, executive compensation, arrogance, globalization, or financial derivatives played any real part here.
Look at the table of contents of this BYTE magazine from 1985. In a nutshell it said the same thing as this article: Functional languages are the great hope for solving the parallel programming problem. Only then the languages were different: Hope, Linda, and Prolog were among them.
My response back then was to get excited about FP. My response now is: Where is the proof? Can anyone name a single instance where a functional paradigm has yielded the best measured performance on a parallel computing problem? In other words, take the best functional programmers in the world, and pair them up with the best tools in existence. Can they actually create something superior, on any problem running on any hardware? This is a very low bar, but until it's demonstrated FP will be confined mostly to the lab.
IMHO the path forward is to treat parallel programming like just another optimization. As we know, the vast majority of your code doesn't need to run fast, and you get most of the performance benefit by optimizing small bits of code that really matter. I suspect the same thing will happen with parallel programming: In a given application only a few areas will benefit much from parallelism, and these tasks will probably be very similar across applications. Graphics rendering, large matrix math, video encoding/decoding, and speech recognition would be examples. People will treat these as special cases, and either develop special-purpose hardware (e.g., GPUs), or libraries that encapsulate the nitty-gritty details. The interesting question to me is what is the best runtime model to support this.
I think people here have a very narrow definition of "software", taken to mean roughly: (a) operating systems, and (b) some productivity apps that business-types use.
Think about software more broadly. Note that console videogame software sales grew 35% in October, compared with October 2007.And this is in spite of a global recession. Can somebody paint a reasonable future scenario where the majority of console games like GTA IV are open-source projects?
My point: You need to adjust your vision of what the software industry is, and where the value is being created. If you want to make money, stay away from commodities like operating systems and word processors.
I think you're right, and the idea of "copyright" in general is headed towards some kind of reform over the long term. Eventually we'll find ourselves in a world where it's not sufficient to have done some valuable work at some point, and then sit around and collect money for the rest of your life.
I can't agree with this view, which I see as wishful thinking. It seems to me that in the fullness of time, the only thing that will be scarce in the information world will be good content. Processing power, storage density, format quality, accessibility: These things are advancing along the technical improvement curve and will soon be close to free. Unfortunately there is no Moore's Law for content. And generally speaking, power and financial reward flow toward the scarcest commodity, not away.
Getting back to the original question, software is content so for the above reasons it should make money over time. The key is to figure out where the profit is to be had. I'll note that GTA IV sold a staggering number of copies; reportedly it was the highest-grossing opening weekend ever in the entertainment industry.
Research by psychologists over the last several decades has strongly supported the idea that most of the variation between individuals is determined by genetics, not environmental factors ("culture", etc.). The old-fashioned 60's idea of gender as a cultural construct isn't believed by very many today. It rapidly falls apart when examined. A couple of good books discussing this are "The Blank Slate" by Steven Pinker, and "The Nurture Assumption" by Judith Rich Harris.
The far more relevant questions have to do with understanding the nature of the genetically-determined differences between men and women (and other genetically distinct groups), and how these differences affect the abilities and preferences of people. Unfortunately these questions are almost taboo for political reasons. Part of the problem is that many people have a hard time understanding what it means to have overlapping broad distributions of traits, with slightly offset means.
I'm not sure what your question is really about. Certainly many companies -- and nearly all that have been around for any length of time -- are profitable. Yahoo is quite profitable, generating over $600M of profit last year. So in that sense I'm not sure what you mean by "speculative gamble". Yes the market moves quickly and companies' fortunes can rise and fall, but it's been that way since capitalism was invented.
With regard to dividend payouts. In a high-growth company, the investors often prefer the company to retain earnings to fund future growth opportunities. When there aren't enough high-value growth opportunities and the cash starts piling up, usually companies will then start paying dividends (Microsoft, for example).
Because of US tax law, it's actually better for most investors if the company uses extra cash to buy back its stock (thus reducing shares outstanding, and increasing the price of the shares that remain), rather than pay it out as a dividend. The former results in a capital gain, which presently in the US is taxed as low as 15%, while the latter counts as ordinary income. Many companies do stock repurchases, sometimes in addition to dividends: Intel and Microsoft for example. This is another perfectly legitimate way to give money back to the investors.
On the contrary, Feynman -- despite having obviously strong mathematical skills -- was I think more driven by physical intuition and experimental work than by mathematics per se. He was quite critical of theorists who pursued physics primarily as a mathematical exercise. Feynman was famously skeptical of string theory for example, which was becoming prominent in the mid-80's toward the end of his life. Overall, one's impression reading Feynman is that he thought of a lot of modern mathematics as window-dressing that got in the way of intuitive understanding. In one of his books he discusses a challenge he had to his mathematician colleagues, in which he said he could answer any true/false question in mathematics -- based on his intuition -- once the basic terms had been defined. I don't know how he actually fared in this challenge, but it clearly shows where his heart lay.
Einstein was also fairly similar in his approach, which was very grounded in intuition and experimental results (sometimes "thought experiments"). His work on general relatively started with the physical principles, and he struggled with the mathematical description for a number of years. Upon describing his work to his mathematician friends (Marcel Grossmann for example), they immediately pointed him toward non-Euclidian (Riemannian) geometry and tensor calculus, subjects he struggled to master and eventually did.
I'm almost as left-wing as they come, but I have to say that nuclear weapons have an unfair bad rap. Compare the number of people killed before and after the invention of the Bomb. According to Wikipedia, WW1 and WW2 killed 20 million and 70 million people, respectively. Since 1945 there has been a lot of chest-thumping by the major powers, but in terms of actual human suffering things are dramatically better. Now we all have big guns pointed at one another, and everything is just fine. On any factual basis you'd have to say the Bomb is the greatest instrument for peace the world has ever seen. Just so long as it doesn't fall into the hands of a rogue individual...
I agree. It's encouraging to think that each of us alive has billions of ancestors over deep time, every last one of which was successful. Think of all the droughts, asteroid impacts, ice ages, fires, plagues, and wars those ancestors lived through and survived. If nothing else, each of us is bred to be very, very scrappy when called upon.
Just as we underplay our own adaptability and will to survive, I think we overplay the power of our technology. In the popular conception, a nuclear war would make humans extinct, or even make all life extinct. Nobody ever justifies these statements. To me they are utter nonsense, and arrogant in the highest degree. In truth we would not be able to kill off even the common mosquito, if all our best minds and economic resources were aimed at it.
It doesn't seem accurate to me to call it "cigar-shaped". If the shape is due to its rotation (which is implied by the Brown et al paper linked from the article), then it would have an oblate spheroid shape like the Earth. I.e., more of a hamburger shape than cigar. It seems a shame to ruin all these Freudian jokes, but facts are facts.
The results here are very interesting. This is different -- and harder -- than the adaptive optics used in ground-based astronomy because the distorting medium is thick, extending all the way to the object being observed. What this implies is that the wavefront distortion isn't uniform across the entire image. So they pick out regions of good (sharp) seeing from each frame, then stitch them together to produce an entire sharp frame. They'll need a fairly fast image processor in those binoculars.
Another option for you would be satellite connectivity, a la HughesNet. It's expensive and the latency isn't great, but the virtue is that it doesn't share any local infrastructure with your other ISPs, except for the power system. Add a generator at your home, and you'd have zero common failure modes and very reliable connectivity overall.
Of course none of this is cheap, so you'll need to decide how much those 7 outages last year are worth to you.
I don't see any sane person adding a computer to his brain for non-medical uses.
Note that technological advances have a way of quickly becoming non-optional. For a certain generation, it was acceptable to not know how to use a computer. What sort of job would you be able to get today if you didn't know how to operate a computer, use the internet, send/receive email, create and print simple documents?
Hypothetically, let's say there is a simple implantable device that will give you all of the internet within milliseconds, just by thinking (Google, Wikipedia, email, etc.).
Here's what I think would happen:
Some people would get the implant, just because they would think it was cool.
Those people would be more productive than the rest of us.
Soon a situation would emerge where the implant was no longer very optional for people who wanted to remain competitive (in the workplace, in creative endeavors, in social opportunities, etc.).
All of this would happen much faster than you might think. (Consider that the last 10 years has seen the internet go from rare plaything to absolutely essential tool for every educated worker.)
It's one thing to say that it takes enormous resources to physically model a real neuron. It's another thing entirely to say that this complexity is necessary to understand its functioning at a logical level.
A transistor is a very complicated device as well, and people write theses and do complex simulations on charge transport, leakage effects, etc. within semiconductor devices. 99% of this detail is irrelevant, however, for understanding what a transistor does within most digital circuits.
It's reasonable to postulate that most of what an actual neuron does is related to biology, rather than computation per se. For starters, a neuron needs to:
obtain energy
physically transport signals without excessive degradation or interference
maintain a proper balance of various ionic species
repair cellular damage
eliminate waste products
etc.
...and presumably many of these details aren't directly relevant to its computational functioning. Put another way, the brain faces tough requirements that an engineered system doesn't face: Self-replication, self-assembly, self-teaching without guidance, biological survival (oxygen, nutrients), etc. and all using a very limited subset of possible materials (proteins).
All that said, I agree with you there is a fascinating question of which particulars of real neural systems need to be captured in order to reproduce their computational properties at an aggregate level. How many "transistors per neuron" as it were. However, part of what Kurzweil is saying here is that, given Moore's Law, it almost doesn't matter what the exact answer to this question is for the purpose of calculating the time to human-level AI.
Let's say that a shuttle could carry up a mile of ribbon. That's 44,000 shuttle launches (22K + counterweight tether.) !!! So, tell me... what will this Space Elevator do that can't be done in the 44,000 shuttle launches NOT COUNTING the dilithium we'll spend getting the counterweight into place?
For this reason, the feasibility studies so far (e.g., the Edwards report) target deployment strategies where first a very low-capacity "pioneer" elevator is established using O(5) conventional rocket launches, then this initial ribbon is reinforced with additional ribbons carried up by climbers on the elevator itself. Over the course of many months and many reinforcements, a high-capacity elevator is built that can bring large-scale payloads to orbit. Obviously an ability to bond these ribbons together at high speed would have to be found -- another thing for the list of Really Hard Engineering Problems needing solution here.
Within this strategy the full counterweight is built up is exactly the same way: Incrementally with mass lifted from the ground, once the pioneer elevator (with its small counterweight) is completed.
The article is misleading: Google's rankings are based entirely on human evaluation. The signals Google uses -- like the number and origin of incoming links to a page, or the specific title and headings the author uses to summarize their content -- are entirely the work of humans trying to make their content useful to readers. Google's algorithms in effect just aggregate this human-created information into a convenient form. (True "algorithmic" search that doesn't use human-generated signals is the problem of general AI, and as-yet unsolved.)
Google also demonstrates that since this information is from a large number of distributed individuals, it's fairly unbiased, reliable, and difficult to game overall. Google can however be gamed (biased) in a narrow specific area through a targeted attack, which I suppose is one issue these competitors are trying to solve. Google's goal of course is to increase the sophistication of the methods it uses in order to filter out these attacks. There is a significant learning curve here, which ultimately is one of Google's advantages over competitors. (Ironically these spam attacks are therefore good for Google, because they keep the search problem hard and allow them to differentiate themselves against competitors.)
It should be recognized that any competitors with hand-edited results pages (Mahalo, etc.) are subject to bias of a different sort: Human bias. If a single editor is creating results pages on specific topics ("abortion" or "sex" or "Taiwan independence" or "George Bush" or "Adolph Hitler" or "Mormonism", etc.), one is naturally concerned about the inherent bias of the editor/author. Wikipedia's approach to this problem is to allow a broad set of people to edit the pages, in the hopes that biases will be detected and ironed out. Since none of these "new" search engines I'm aware of allow this type of openness, it seems to me they run a real risk of just being crappier versions of Wikipedia.
Sir Clarke will no doubt be remembered by many as a futurist and inventor, or at least foreteller, of some important technological developments of the 20th century. While these are important contributions, I believe his lasting impact is much greater.
The essential genius of Clarke's work is the sense of wonder and mystery and human potential that permeates it. His writing captures perfectly the deep awe and curiosity and wonder I feel when I look into the night sky and grope feebly toward comprehension. Scientists and spiritualists alike are in truth motivated in the end by exactly the same thing: The deep mystery at the core of existence, that black hole in our limited understanding which we perceive but cannot see. Clarke's writing acknowledges the deep mysteries and their power over us, and does not aim to trivialize with ready explanations or tidy conclusions. Many often criticize the "confusing" endings of 2001 or Childhood's End or his other works, but they could be no other way.
And yet although Clarke sees us as fundamentally limited creatures incapable of full comprehension, his message is also relentlessly hopeful. The deep mystery of things is not something for us to worship or fear or ignore, but rather to seek out and face head-on with vigor and modesty and self-awareness. Not with intent to conquer, but with intent to appreciate and learn. This is the greatest hope for our future.
Sir Clarke was deep and true in his understanding. His work will be as relevant a millenium from now as it is today.
Probably the same way the blank media levy is collected/distributed: lump sums given out to the songwriters' and musicians' guilds, which is then distributed by the guild on basis of need. Quite a fair way to do things, really, and one that the majority of Canadian musicians support wholeheartedly.
Distributing on the basis of "need" is a poor way to do it from the standpoint of incentives. Why would an artist bother to create good music when they could just sit around being "needy"?
A much better approach would be to emulate the relationships performance rights organizations (ASCAP and BMI) have with radio stations. The money that is collected from their blanket licenses (http://www.ascap.com/licensing/radio/radiofaq.html) is distributed to artists based on how many times their songs are played. In other words, you should monitor the pattern of downloading activity and distribute the income on that basis.
I actually own an HD-DVD player, and I want the whole damned thing to die.
My challenge to hardware engineers everywhere: Just give me a friggin "Play" button. I don't want to watch previews. I don't want to watch cute animated menus with everything in different places because no interface guidelines exist. PLEASE DO NOT give UI designers a Turing-complete language to play with. And if you find yourself producing a video playback device with more than 70 buttons on the remote control, please reflect upon the fact that you may have taken a wrong turn somewhere in the design process.
DVD is bad enough, but the HD-DVD disks I've used elevate the problems to a whole new level of Hell. I have a Ph.D and I can't navigate these things. Watching a movie last night it took me literally 5 minutes to figure out how to turn subtitles on and off. The problem is that all of these things are being engineered by Japanese companies, and for all their strengths one thing about the Japanese is that they produce crappy software and user interfaces. There, I said it.
I hope that internet-delivered video will win. If not from the studios, then from some brave soul who has navigated the HD-DVD nightmare and ripped a clean copy into BitTorrent that I can just friggin PLAY.
I think that will change when Ma & Pa Kettle waddle into the Big-Box store to get their new Digital-TV and the Blue-Shirts up-sell them to a 42" LCD, surround sound and Bluray player package so they can watch Raw and Smackdown in all it's glory!
The point is, the 42" LCD and surround speakers are major improvements to the viewing experience. HD-DVD (or Bluray) not so much IMHO. (This is based on actual experience: I have a 42" HDTV, surround speakers, and an HD-DVD player.) Some HD-DVD content is very good -- for example BBC's "Planet Earth" documentary -- but most movies don't look much better than DVD. Either the original source material isn't sharp enough to make a difference, or the studios are cutting corners on HD-DVD authoring. As HD becomes more prevalent this may change, but for now the visual difference is subtle at best. Broadcast HDTV however can be stunning, but again IMHO these disks aren't delivering a visual experience of that quality.
When the price of the units comes down to $150, it will make sense to buy one when you're in the market for a new DVD player. But shell out $300+ for one of these things? As an owner of one I can say you'd be better off spending that money on a larger screen, or better speakers.
I propose that we, the/. community, establish a vacation fund for New Zealand physicists.
I strongly suspect this author is not a physicist, but rather a computer scientist. Three points:
His email address is @acm.org
A computer science organization is referenced on the cover of the paper
He says things that very few physicists would say, like "quantum entanglement implies faster-than-light communication"
So he is a programmer, and in fact I believe he is the one who programmed our simulation. I suggest that someone kidnap him and demand to be instantiated in his higher level reality. Then you can help the rest of us escape too.
Slashdot Mirror
I was monkeying around with a C64 emulator the other day, and it struck me how bad those old OSes were. I do have some nostalgia for these things, but more for the times they represented in my life than because I miss the hardware and software. In truth they were mostly cobbled-together messes.
BeOS is the only one I truly miss, and that is because it had something none of the current OSes have: Low user latency. With the current crop of OSes we take it for granted that:
What I miss about BeOS was the whole design aesthetic of putting the user first, never blocking user input, and making the common use cases fast.
As an experiment in grad school, for about 6 months I completely switched over to Dvorak. It wasn't about becoming a faster typist; I'm a pretty capable typist on QWERTY (> 100 wpm), and in a practical sense I find that thinking of what to type is usually the rate-limiting factor. I had two goals: (a) See whether and how quickly my brain could retrain, and (b) See whether there was any impact on some mild pain in my wrists (I was doing a lot of typing).
As others have pointed out, with keyboard remapping it's easy to get Dvorak on any computer you might happen to use. So in a practical sense there is no longer any lock-in of that type.
My experience:
Ultimately I switched back to QWERTY for the keyboard shortcut issue noted above. Dvorak didn't have nearly as much impact on my wrist pain as taking breaks from work did.
My conclusion is that QWERTY vs Dvorak is largely irrelevant. There may have been more relevance before computers, when a lot of typing was done as dictation or copying. However most people today type as they're composing the words, and I strongly suspect brain speed is the practical limiting factor, not keyboard layout.
Well, the short answer is that the world needs more than one financial instrument for the same reason it needs more than one programming language, operating system, microprocessor variety, animal species, type of food, and so on. It's because diversity is inherently good. (Except when it isn't, which is pretty rare. Sometimes the benefit of standardization trumps diversity, e.g., electrical plugs and rail gauges.) Diversity creates competition, and opens up room for innovation. It also makes a system more, not less, resilient to catastrophe.
If you're looking for something to blame for your vanishing 401(k), here is my starter list:
I doubt that financial product complexity is on the list. I'm also skeptical that greed, executive compensation, arrogance, globalization, or financial derivatives played any real part here.
Look at the table of contents of this BYTE magazine from 1985. In a nutshell it said the same thing as this article: Functional languages are the great hope for solving the parallel programming problem. Only then the languages were different: Hope, Linda, and Prolog were among them.
My response back then was to get excited about FP. My response now is: Where is the proof? Can anyone name a single instance where a functional paradigm has yielded the best measured performance on a parallel computing problem? In other words, take the best functional programmers in the world, and pair them up with the best tools in existence. Can they actually create something superior, on any problem running on any hardware? This is a very low bar, but until it's demonstrated FP will be confined mostly to the lab.
IMHO the path forward is to treat parallel programming like just another optimization. As we know, the vast majority of your code doesn't need to run fast, and you get most of the performance benefit by optimizing small bits of code that really matter. I suspect the same thing will happen with parallel programming: In a given application only a few areas will benefit much from parallelism, and these tasks will probably be very similar across applications. Graphics rendering, large matrix math, video encoding/decoding, and speech recognition would be examples. People will treat these as special cases, and either develop special-purpose hardware (e.g., GPUs), or libraries that encapsulate the nitty-gritty details. The interesting question to me is what is the best runtime model to support this.
I think people here have a very narrow definition of "software", taken to mean roughly: (a) operating systems, and (b) some productivity apps that business-types use.
Think about software more broadly. Note that console videogame software sales grew 35% in October, compared with October 2007. And this is in spite of a global recession. Can somebody paint a reasonable future scenario where the majority of console games like GTA IV are open-source projects?
My point: You need to adjust your vision of what the software industry is, and where the value is being created. If you want to make money, stay away from commodities like operating systems and word processors.
I think you're right, and the idea of "copyright" in general is headed towards some kind of reform over the long term. Eventually we'll find ourselves in a world where it's not sufficient to have done some valuable work at some point, and then sit around and collect money for the rest of your life.
I can't agree with this view, which I see as wishful thinking. It seems to me that in the fullness of time, the only thing that will be scarce in the information world will be good content. Processing power, storage density, format quality, accessibility: These things are advancing along the technical improvement curve and will soon be close to free. Unfortunately there is no Moore's Law for content. And generally speaking, power and financial reward flow toward the scarcest commodity, not away.
Getting back to the original question, software is content so for the above reasons it should make money over time. The key is to figure out where the profit is to be had. I'll note that GTA IV sold a staggering number of copies; reportedly it was the highest-grossing opening weekend ever in the entertainment industry.
Research by psychologists over the last several decades has strongly supported the idea that most of the variation between individuals is determined by genetics, not environmental factors ("culture", etc.). The old-fashioned 60's idea of gender as a cultural construct isn't believed by very many today. It rapidly falls apart when examined. A couple of good books discussing this are "The Blank Slate" by Steven Pinker, and "The Nurture Assumption" by Judith Rich Harris.
The far more relevant questions have to do with understanding the nature of the genetically-determined differences between men and women (and other genetically distinct groups), and how these differences affect the abilities and preferences of people. Unfortunately these questions are almost taboo for political reasons. Part of the problem is that many people have a hard time understanding what it means to have overlapping broad distributions of traits, with slightly offset means.
I'm not sure what your question is really about. Certainly many companies -- and nearly all that have been around for any length of time -- are profitable. Yahoo is quite profitable, generating over $600M of profit last year. So in that sense I'm not sure what you mean by "speculative gamble". Yes the market moves quickly and companies' fortunes can rise and fall, but it's been that way since capitalism was invented.
With regard to dividend payouts. In a high-growth company, the investors often prefer the company to retain earnings to fund future growth opportunities. When there aren't enough high-value growth opportunities and the cash starts piling up, usually companies will then start paying dividends (Microsoft, for example).
Because of US tax law, it's actually better for most investors if the company uses extra cash to buy back its stock (thus reducing shares outstanding, and increasing the price of the shares that remain), rather than pay it out as a dividend. The former results in a capital gain, which presently in the US is taxed as low as 15%, while the latter counts as ordinary income. Many companies do stock repurchases, sometimes in addition to dividends: Intel and Microsoft for example. This is another perfectly legitimate way to give money back to the investors.
On the contrary, Feynman -- despite having obviously strong mathematical skills -- was I think more driven by physical intuition and experimental work than by mathematics per se. He was quite critical of theorists who pursued physics primarily as a mathematical exercise. Feynman was famously skeptical of string theory for example, which was becoming prominent in the mid-80's toward the end of his life. Overall, one's impression reading Feynman is that he thought of a lot of modern mathematics as window-dressing that got in the way of intuitive understanding. In one of his books he discusses a challenge he had to his mathematician colleagues, in which he said he could answer any true/false question in mathematics -- based on his intuition -- once the basic terms had been defined. I don't know how he actually fared in this challenge, but it clearly shows where his heart lay.
Einstein was also fairly similar in his approach, which was very grounded in intuition and experimental results (sometimes "thought experiments"). His work on general relatively started with the physical principles, and he struggled with the mathematical description for a number of years. Upon describing his work to his mathematician friends (Marcel Grossmann for example), they immediately pointed him toward non-Euclidian (Riemannian) geometry and tensor calculus, subjects he struggled to master and eventually did.
I'm almost as left-wing as they come, but I have to say that nuclear weapons have an unfair bad rap. Compare the number of people killed before and after the invention of the Bomb. According to Wikipedia, WW1 and WW2 killed 20 million and 70 million people, respectively. Since 1945 there has been a lot of chest-thumping by the major powers, but in terms of actual human suffering things are dramatically better. Now we all have big guns pointed at one another, and everything is just fine. On any factual basis you'd have to say the Bomb is the greatest instrument for peace the world has ever seen. Just so long as it doesn't fall into the hands of a rogue individual...
I agree. It's encouraging to think that each of us alive has billions of ancestors over deep time, every last one of which was successful. Think of all the droughts, asteroid impacts, ice ages, fires, plagues, and wars those ancestors lived through and survived. If nothing else, each of us is bred to be very, very scrappy when called upon.
Just as we underplay our own adaptability and will to survive, I think we overplay the power of our technology. In the popular conception, a nuclear war would make humans extinct, or even make all life extinct. Nobody ever justifies these statements. To me they are utter nonsense, and arrogant in the highest degree. In truth we would not be able to kill off even the common mosquito, if all our best minds and economic resources were aimed at it.
It doesn't seem accurate to me to call it "cigar-shaped". If the shape is due to its rotation (which is implied by the Brown et al paper linked from the article), then it would have an oblate spheroid shape like the Earth. I.e., more of a hamburger shape than cigar. It seems a shame to ruin all these Freudian jokes, but facts are facts.
The article has a broken link to the original technical presentation. Try this: http://www.iol.umd.edu/Presentations/slideshow.php?id=54
The results here are very interesting. This is different -- and harder -- than the adaptive optics used in ground-based astronomy because the distorting medium is thick, extending all the way to the object being observed. What this implies is that the wavefront distortion isn't uniform across the entire image. So they pick out regions of good (sharp) seeing from each frame, then stitch them together to produce an entire sharp frame. They'll need a fairly fast image processor in those binoculars.
Another option for you would be satellite connectivity, a la HughesNet. It's expensive and the latency isn't great, but the virtue is that it doesn't share any local infrastructure with your other ISPs, except for the power system. Add a generator at your home, and you'd have zero common failure modes and very reliable connectivity overall.
Of course none of this is cheap, so you'll need to decide how much those 7 outages last year are worth to you.
I have one of these posters. It looks pretty cool, and is mathematics-related:
http://www.vcalc.net/cu.htm#Curta-PosterI don't see any sane person adding a computer to his brain for non-medical uses.
Note that technological advances have a way of quickly becoming non-optional. For a certain generation, it was acceptable to not know how to use a computer. What sort of job would you be able to get today if you didn't know how to operate a computer, use the internet, send/receive email, create and print simple documents?
Hypothetically, let's say there is a simple implantable device that will give you all of the internet within milliseconds, just by thinking (Google, Wikipedia, email, etc.).
Here's what I think would happen:
It's one thing to say that it takes enormous resources to physically model a real neuron. It's another thing entirely to say that this complexity is necessary to understand its functioning at a logical level.
A transistor is a very complicated device as well, and people write theses and do complex simulations on charge transport, leakage effects, etc. within semiconductor devices. 99% of this detail is irrelevant, however, for understanding what a transistor does within most digital circuits.
It's reasonable to postulate that most of what an actual neuron does is related to biology, rather than computation per se. For starters, a neuron needs to:
...and presumably many of these details aren't directly relevant to its computational functioning. Put another way, the brain faces tough requirements that an engineered system doesn't face: Self-replication, self-assembly, self-teaching without guidance, biological survival (oxygen, nutrients), etc. and all using a very limited subset of possible materials (proteins).
All that said, I agree with you there is a fascinating question of which particulars of real neural systems need to be captured in order to reproduce their computational properties at an aggregate level. How many "transistors per neuron" as it were. However, part of what Kurzweil is saying here is that, given Moore's Law, it almost doesn't matter what the exact answer to this question is for the purpose of calculating the time to human-level AI.
Let's say that a shuttle could carry up a mile of ribbon. That's 44,000 shuttle launches (22K + counterweight tether.) !!! So, tell me... what will this Space Elevator do that can't be done in the 44,000 shuttle launches NOT COUNTING the dilithium we'll spend getting the counterweight into place?
For this reason, the feasibility studies so far (e.g., the Edwards report) target deployment strategies where first a very low-capacity "pioneer" elevator is established using O(5) conventional rocket launches, then this initial ribbon is reinforced with additional ribbons carried up by climbers on the elevator itself. Over the course of many months and many reinforcements, a high-capacity elevator is built that can bring large-scale payloads to orbit. Obviously an ability to bond these ribbons together at high speed would have to be found -- another thing for the list of Really Hard Engineering Problems needing solution here.
Within this strategy the full counterweight is built up is exactly the same way: Incrementally with mass lifted from the ground, once the pioneer elevator (with its small counterweight) is completed.
The article is misleading: Google's rankings are based entirely on human evaluation. The signals Google uses -- like the number and origin of incoming links to a page, or the specific title and headings the author uses to summarize their content -- are entirely the work of humans trying to make their content useful to readers. Google's algorithms in effect just aggregate this human-created information into a convenient form. (True "algorithmic" search that doesn't use human-generated signals is the problem of general AI, and as-yet unsolved.)
Google also demonstrates that since this information is from a large number of distributed individuals, it's fairly unbiased, reliable, and difficult to game overall. Google can however be gamed (biased) in a narrow specific area through a targeted attack, which I suppose is one issue these competitors are trying to solve. Google's goal of course is to increase the sophistication of the methods it uses in order to filter out these attacks. There is a significant learning curve here, which ultimately is one of Google's advantages over competitors. (Ironically these spam attacks are therefore good for Google, because they keep the search problem hard and allow them to differentiate themselves against competitors.)
It should be recognized that any competitors with hand-edited results pages (Mahalo, etc.) are subject to bias of a different sort: Human bias. If a single editor is creating results pages on specific topics ("abortion" or "sex" or "Taiwan independence" or "George Bush" or "Adolph Hitler" or "Mormonism", etc.), one is naturally concerned about the inherent bias of the editor/author. Wikipedia's approach to this problem is to allow a broad set of people to edit the pages, in the hopes that biases will be detected and ironed out. Since none of these "new" search engines I'm aware of allow this type of openness, it seems to me they run a real risk of just being crappier versions of Wikipedia.
Sir Clarke will no doubt be remembered by many as a futurist and inventor, or at least foreteller, of some important technological developments of the 20th century. While these are important contributions, I believe his lasting impact is much greater.
The essential genius of Clarke's work is the sense of wonder and mystery and human potential that permeates it. His writing captures perfectly the deep awe and curiosity and wonder I feel when I look into the night sky and grope feebly toward comprehension. Scientists and spiritualists alike are in truth motivated in the end by exactly the same thing: The deep mystery at the core of existence, that black hole in our limited understanding which we perceive but cannot see. Clarke's writing acknowledges the deep mysteries and their power over us, and does not aim to trivialize with ready explanations or tidy conclusions. Many often criticize the "confusing" endings of 2001 or Childhood's End or his other works, but they could be no other way.
And yet although Clarke sees us as fundamentally limited creatures incapable of full comprehension, his message is also relentlessly hopeful. The deep mystery of things is not something for us to worship or fear or ignore, but rather to seek out and face head-on with vigor and modesty and self-awareness. Not with intent to conquer, but with intent to appreciate and learn. This is the greatest hope for our future.
Sir Clarke was deep and true in his understanding. His work will be as relevant a millenium from now as it is today.
Probably the same way the blank media levy is collected/distributed: lump sums given out to the songwriters' and musicians' guilds, which is then distributed by the guild on basis of need. Quite a fair way to do things, really, and one that the majority of Canadian musicians support wholeheartedly.
Distributing on the basis of "need" is a poor way to do it from the standpoint of incentives. Why would an artist bother to create good music when they could just sit around being "needy"?
A much better approach would be to emulate the relationships performance rights organizations (ASCAP and BMI) have with radio stations. The money that is collected from their blanket licenses (http://www.ascap.com/licensing/radio/radiofaq.html) is distributed to artists based on how many times their songs are played. In other words, you should monitor the pattern of downloading activity and distribute the income on that basis.
...is "out-house" data centers. Powered entirely by human waste. Very green, very modern, it's recycling for the new millennium.
I actually own an HD-DVD player, and I want the whole damned thing to die.
My challenge to hardware engineers everywhere: Just give me a friggin "Play" button. I don't want to watch previews. I don't want to watch cute animated menus with everything in different places because no interface guidelines exist. PLEASE DO NOT give UI designers a Turing-complete language to play with. And if you find yourself producing a video playback device with more than 70 buttons on the remote control, please reflect upon the fact that you may have taken a wrong turn somewhere in the design process.
DVD is bad enough, but the HD-DVD disks I've used elevate the problems to a whole new level of Hell. I have a Ph.D and I can't navigate these things. Watching a movie last night it took me literally 5 minutes to figure out how to turn subtitles on and off. The problem is that all of these things are being engineered by Japanese companies, and for all their strengths one thing about the Japanese is that they produce crappy software and user interfaces. There, I said it.
I hope that internet-delivered video will win. If not from the studios, then from some brave soul who has navigated the HD-DVD nightmare and ripped a clean copy into BitTorrent that I can just friggin PLAY.
I think that will change when Ma & Pa Kettle waddle into the Big-Box store to get their new Digital-TV and the Blue-Shirts up-sell them to a 42" LCD, surround sound and Bluray player package so they can watch Raw and Smackdown in all it's glory!
The point is, the 42" LCD and surround speakers are major improvements to the viewing experience. HD-DVD (or Bluray) not so much IMHO. (This is based on actual experience: I have a 42" HDTV, surround speakers, and an HD-DVD player.) Some HD-DVD content is very good -- for example BBC's "Planet Earth" documentary -- but most movies don't look much better than DVD. Either the original source material isn't sharp enough to make a difference, or the studios are cutting corners on HD-DVD authoring. As HD becomes more prevalent this may change, but for now the visual difference is subtle at best. Broadcast HDTV however can be stunning, but again IMHO these disks aren't delivering a visual experience of that quality.
When the price of the units comes down to $150, it will make sense to buy one when you're in the market for a new DVD player. But shell out $300+ for one of these things? As an owner of one I can say you'd be better off spending that money on a larger screen, or better speakers.
I propose that we, the /. community, establish a vacation fund for New Zealand physicists.
I strongly suspect this author is not a physicist, but rather a computer scientist. Three points:
So he is a programmer, and in fact I believe he is the one who programmed our simulation. I suggest that someone kidnap him and demand to be instantiated in his higher level reality. Then you can help the rest of us escape too.