As a side note, many scenes in those stereoscopic toys (disk with ~dozen photos) that I've seen had very deep focus... IMHO it makes the whole scene, paradoxically, very flat. Yes, there is "depth" of course - but feels non-gradual, like several backgrounds in old SNES platformers.
What is the purpose of this post? What does it even mean? What is the purpose of posting a link to a nebulous summary of a highly suggestive report on an extremely politically charged subject on a site that bills itself "News for Nerds"?
But yes, it's true: I don't know MS-SQL, and nobody else at EDA does either. So we were faced with a choice: find a few people who did know it, pay 'em a bunch of donated money to write a formal report behind closed doors, or do a public review and exam even if that means exposing any mistakes we make, knowing they'll be caught pretty damn quick.
False dichotomy, based on the unproven assumption that there was anything nefarious going on that required urgent exposition.
With all due respect to an eminent and brilliant physicist, Freeman Dyson is not a climatologist.
But Dyson's argument in fact questions the methods and claims of the science of climatology. If his argument is valid, then it does not matter that he is not a climatologist - it makes Dyson's criticisms more urgent, not less.
Every discussion around HTML and related W3C standards always seems to end up in a blamefest. Microsoft is to blame for their poor standards compliance, lazy web developers are to blame for sloppy code, Al Gore is to blame for inventing the Internet.
All of that is true. But I have come to believe that perhaps the blame lies primarily with the standards themselves. They are just not very good.
I know this is not a popular opinion. Let me qualify it and try to explain briefly what I mean. There is of course a lot of theoretical and historical background to consider, but frankly it is a waste of time to drag all that into a Slashdot comment.
The first problem is with HTML. HTML abstracts at the wrong level. It should be a presentation language, not a structural markup language. There is no need for HTML as a structural markup language and frankly I am baffled by the religious zeal with which some people defend this notion. As a structural markup language, HTML is very poor. Structural markup is most useful for well-defined, domain-specific applications. That is not what HTML is used for and this causes numerous problems: ill-defined rendering behavior, poor querying and indexing abilities, poor feature set, relatively slow performance, not to mention poor reusability.
The second problem is with CSS. Although at its core a good idea, it is poorly implemented, with a pointlessly weird, C-inspired syntax. It is too feeble to express presentational structure and lacks a method to express generalized context-dependent relationships. The selector language is so baroque that it is poorly understood by authors and implementors alike. Most importantly, CSS simply does not solve many common layout and styling problems, except at the most trivial level. Efforts to address this have mostly just made CSS more complicated rather than more powerful.
The third problem is with Javascript. The language itself is not bad, but it exists in an environment that is so primitive and crude, that often the easiest way to accomplish anything is still to just stuff precalculated strings into a node's innerHTML. The web is littered with the corpses of Javascript libraries to provide simple services like data binding, templating, input validation and widget sets. None of which build on eachother, because there are no tools to enable this kind of workflow, and most of which fail on either correctness, performance or standards-compliance.
Why are there so many problems with these standards? Is it normal for standards to be so problematic? It is certainly true that numerous standards failed. But on the other hand, many other standards succeeded. PostScript and PDF are very successful standards that have been implemented dozens of times with minimal interoperability issues. The same goes for countless file format standards, such as GIF, PNG, JPG and ZIP, or standards such as ASCII or Unicode.
Of course the comparison between HTML (and all related tech) and, say, GIF, is not valid. In the case of HTML there are many reasons, some socio-economic, which have brought us to the point where we are today. But despite that, I believe it is possible to identify 2 key issues with the W3C family of tech:
The wrong abstraction level. The W3C people have been chasing a nebulous vision of a "semantic web" which is accessible for everyone on any kind of device. This has resulted in intentionally vague, abstract specifications. But people who have not done a lot of work building GUIs for production systems do not really understand how to abstract layout and presentation. This was a key failing in the original Java GUI toolkit called AWT (Abstract Window Toolkit).
The wrong people. The HTML/CSS/Javascript/XSL standards have been developed by people who are primarily interested in information technology and theory. They have little understanding of and less experience with graphic design and application front-end development. They really do not understand what distinguishes good from bad in this area
It's probably best if you start off learning what SWFs are and how the Flash "IDE" works before you start looking at the alternatives. It's a pretty complex system with heavy emphasis on graphics and animation and a large amount of cruft. You can trial Flash for 30 days, after that you can look at tools like FDT, MTASC, SWFmill and FlashDevelop. See osflash.org for a variety of open source tools around the Flash platform (like MTASC and SWFmill).
FlashDevelop and the FDT plugin for Eclipse, to name just two, both provide a very good development environment, in many ways better (from a developers perspective) than the Adobe IDE. Of course the Adobe IDE also doubles as a design environment, and in that role it has virtually no competition. But this has traditionally been the case with design tools. Because designers heavily invest in a specific tool (like Photoshop, InDesign or Quark), and those skills are harder to transfer between different tools than the skills of a developer.
Not really. The SWF file format is open, and there are quite a number of quality open-source Flash development tools. There is no open source Flash runtime which is as good as the Adobe runtime, but then there is no Java runtime which is as good as Sun's runtime either.
Most of the commentary on this topic is shit, fueled by ignorance and unthinking dogmatism.
Over the past half decade or so, Microsoft has been developing arguably the most comprehensive and coherent development platform ever on the planet, viz..NET. Many people like to denounce.NET as "Java copied badly" or point out how poorly Windows Forms compares to what's available for GNOME and MacOS X.
This kind of argument is completely besides the point. While some parts of.NET are not be as good as other offerings on the market, as a whole there is nothing which compares to it..NET brings everything under one roof and eliminates entire classes of "glue" and "can't get there from here" problems.
(Yes, we are all software developers and enthusiasts. We all know the joys of loosely coupled systems and the evils of integration. I'm realy not interested in a generic discussion on that. In practice all good things have costs and all bad things have benefits and.NET in most cases does The Right Thing. If you haven't worked with.NET yet, just try it and come with specifics. Don't come arguing on abstract principles please.)
With.NET 3.0 and WPF, a brand new UI subsystem has been added to the mix, which in terms of raw capability rivals anything out there.... Christ, that sounds like a commercial. But it's true. You've all seen the demos of movies projected onto flying 3D surfaces etcetera, and this might have left you with the impression that there is little substance to the technology apart from fizz and sparkle.
That would be a very wrong impression..NET and WPF form the foundation for the next generation of Windows applications and Silverlight brings parts of this technology to the web. Thus, while Silverlight may falter, as some of you have been suggesting, the underlying technology certainly will not be going anywhere anytime soon.
Therefore to suggest that Miguel or "we" could or even should be developing an "alternative to Silverlight" is absolute nonsense and indicates an utter blindness for the bigger picture.
The whole point of.NET is that it provides a clean and sane means of unifying traditionally separate realms of development. With.NET and Silverlight, it is slowly becoming possible to leverage the same skills and code on the Web (both server side and client side), the desktop, games consoles, set top boxes, PDAs and Mobile phones.
Even if you develop something that's significantly better than.NET for some specific task/domain, it would have to be several orders of magnitudes better before the marginal benefit offsets the costs of not being able to ride the slipstream of the Microsoft/.NET juggernaut.
Microsoft has been busy rewriting their entire crufty codebase to a modern, unified platform. We are still arguing over widget sets and the relative merits of the GNOME file selector dialog vs. the KDE one. Wake up people.
"Those two headlights heading towards me at high speed could be a bus that will kill me, but they could be two motorcyles that will pass harmlessly on either side of me. I think I'll stand here until I know for sure."
Were it so simple. In your example there is no benefit to remaining where you are and no penalty for stepping aside. In the real world the benefit of the status quo is huge and the penalty for change enormous.
Look, the only reason why Intel and HP keep sinking money into this dog is because they've already spent incredible amounts on it and have been telling everyone, for about a decade now, that it's the Future of Enterprise Computing.
They seem unable or unwilling to face up to the fact that the Itanium is a complete albatross, and there's really nothing admirable about that.
Again, garbage. Progress comes from the 100 corrobarating experiments that may have been suggested by the original anomoly. It does NOT come from throwing out existing theories because of a single one off result. The 'isolated hero' model of science is just hollywood bunkum.
I was referring to Popper's falsification principle. A thousand white swans might support the hypothesis that all swans are white, but just a single black swan disproves it.
Now the problem with the global warming hypothesis is that it has become increasingly resistant to falsification. We are finding a number of black or at least grey swans;
Measurements from sattelites and weatherballoons fail to show the expected increase in upper air temperature.
Ice core measurements show that an increase in CO2 has historically never preceded an increase in temperature, but always trailed it. As such CO2 can be ruled out as a causal agent (although it is likely to play a sustaining role).
NASA readings link increased solar radiation to unprecedented polar ice cap melting on Mars and Pluto as well as Earth; this would certainly suggest a strong non-human component to any real or perceived climate change here on Earth.
CO2 and Methane constitute a tiny, tiny fraction of the Earth's atmosphere, on the order of 0.1% to 0.01%. So far nobody has conclusively demonstrated how fluctuations in such a tiny quantity might effect the huge predicted changes in climate.
Taken together, these observations raise serious questions about the degree to which we understand global warming and the degree to which we can or should prevent it.
That the Earth is getting warmer on average is probably a given; everyone is familiar with the fact that the temperature in big cities can be up to a full degree or two higher than in the countryside, and combined with the fact that the planet is becoming increasingly urbanized, it stands to reason that the Earth's temperature will have increased by some amount over the last 50 or 100 years.
The real question is whether this is a problem. And to the extent that it is indeed a problem, what is the best solution? It costs billions and billions to effect the few percent CO2 emission reduction as required by Kyoto; but the evidence for CO2 being a causal agent in global warming is very thin; and the ultimate effect of this reduction in terms of e.g. environmental damage or sea level rise is unpredictable and perhaps negligible. Then considering all this, and considering that all the money we sink into emission reductions can no longer be spent on projects that have a directly measurable effect, such as better public transport, or wildlife conservation, or water management; is this really a windmill we want to be tilting at?
There's about 2 million more out there, thanks to a strong consensus by people who do this for a living.
Again, science is not a consensus discipline. For example, the strong consensus in the '70s was that we were about to enter a new ice age. Nothing of the sort seems to be the case today. In a very real sense, scientific progress comes about not from the 100 corroborating experiments but from the single anomalous one.
More importantly your link doesn't address the fundamental questions: to what extent is the climate changing, what kinds of effects will this have, to what extent can or should we prevent it?
The original poster, being a fanatic, doesn't entertain any of these questions. For a fanatic, it's only important to know that a) the sky is falling, b) we are to blame, and c) they are morally superior to everyone who doesn't see it that way.
Whereas for normal people, the question is one of compromise: to what extent is climate change actually happening, to what extent can we manage or control the risks of climate change, and to what extent do we sacrifice other desirable goals in order to curb the risk.
Look, you're a fanatic and a jackass with the good fortune to live in an era where fanatics and jackasses rule the world. Which is why there is really no point in arguing with you or upholding the pretense of a cultured back-and-forth; you'll just yell, snark and cajole your way back to the consensus reality you find so comforting. Still one remains hopeful that even a boor like you might recognize the bright flame of reason as a liberation from, rather than as a threat to his miserable existence, and it is in that spirit that I shall leave you with the following:
On average, CO2 and Methane make up less than 0.1-0.01% of the Earth's atmosphere, combined. To be sure, this is a very, very small amount. Kyoto strives to reduce CO2 emissions by 5%. The approximate cost of this reduction, which is miniscule and wholly within the margin of error on a global scale, amounts to tens of billions of dollars which can no longer be invested in education, medicine or social benefits. Is this good governance or just fanaticism?
NASA has presented evidence of increased solar radiation since the '70s and postulated a link to climate change -- not just on Earth, but on Mars and Pluto as well, where measurements have shown unprecedented polar ice cap shrinkage over the past half decade. Why is this fact never mentioned in media reporting on the phenomenon on climate change?
To the extent that the climate is indeed changing, regardless of the cause, we are extremely poor at predicting the local effects of this change and we do not know to what extent we are capable of controlling it. Spending billions trying to avert a disaster, the magnitude of which is unknown and may be very small indeed, that may happen sometime in the next century or perhaps not at all, that we may not be able to avert in the first place, while every winter an increasing number of old people in the developed world are freezing to death because they can't pay their energy bills and levees are being underfunded, is that good governance or just a really expensive form of penitence for the perpetual guilt-trip of the fanatic?
You are asking us to believe that temperature data from two isoloated stations should contradict the overall trend from hundreds of thousands of stations across the entire planet over hundreds of years.
At least he provided data. You provide only babble.
This isn't crazy hippies crying about killing the poor flowers somewhere. This is a near perfect consensus of scientists across dozens of different disciplines on a global scale.
Science is not a consensus discipline. At the same time scientists are not immune to groupthink. The "scientific consensus" has included howlers such as the subhumanity of negros and the notion that every boy wants to fuck his mother.
Just uneducated, dishonest conservatives who see global climate change as a threat to their monied way of life, and who are willing to risk making a world of Katrina-like events commonplace to protect that way of life.
The root of the matter is that you and your cronies fundamentally oppose the wealth and affluence (and, yes, the inherent waste) of the society we live in, and will use any excuse at your disposal to curb its further growth, be it global warming, overpopulation, the dawning of a new ice age, nuclear scares, peak oil or worried aliens.
Putting linux on machines designed to prevent that very thing is like a game of football for geeks. It requires skill (is not too easy), but has been and probably can be done (is not too hard). The rules are those of logic and electronics. The goal is clear, and there is feedback along the way as you (carefully arrange to) see evidence of the system running your code further and further along in the boot process
The same can be said of masturbation. Which this is.
It's true that planning is hard and software is frequently late, but the allegation here is not that Microsoft is bad at planning. The allegation is that they purposefully and intentionally announce unrealistic schedules in order to string along their customers. By announcing that the next release of Windows is "just around the corner" customers are dissuaded from evaluating the offerings of the competition.
Proposals like these are stupid and idiotic because half the time it simply ISN'T CLEAR what a piece of software should be doing. We simply don't have formally rigorous and perfect descriptions to meet real-world demands, and for the most part we CAN'T have them. If we had, we could use those instead of writing computer programs. For the time being, however, computer programs are the best descriptions we have.
Furthermore, the "intelligence" that a human uses in using a vacuum cleaner is not conscious thought, but comes from "brute force" computation in the brain.
This is a gross misappropriation of terms, because it is highly debatable whether the processes in the brain can be reduced to computational processes. The "brain=hardware, mind=software" metaphor befuddles as much as it explains.
Intelligence should be about end results. If someone else solved a puzzle more quickly than me, it would seem odd to regard him as unintelligent just because he used a different method to me.
This touches on the whole intentionalism debate. Your specific example is flawed because we really do not know what tasks require intelligence. We tend to think intelligence is a prerequisite for solving puzzles because we've only ever seen intelligent beings solve puzzles. But the singular most important discovery made through AI research has been that for many tasks, you don't need intelligence.
If in x years time we have proper speech recognition, but it comes primarily through advancements in hardware rather than more clever algorithms, will speech recognition join chess playing as examples of AI which aren't really intelligent?
If it turns out that you don't need intelligence to recognize speech, sure.
And once this is achieved, people like you will be saying it's not AI, because it doesn't come in the form of a humanoid robot which turns the steering wheel and pushes the pedals!
It's just a driving machine. It seems quite probable that you don't need intelligence to build one. But so far we can't even manage that.
I don't mean to be a sophist. I just think the extentionalist approach to AI, notwithstanding its obvious successes, has more or less exhausted its potential. Because underlying its hard-nosed empiricism is the very magical belief that if only we can create accurate enough simulations, the real thing will spontaneously emerge. Essentially it amounts to cargo cultism.
If it's more efficient to have several dedicated devices rather than an all-in-one, why does this mean AI has failed?
The failure is that instead of building a device with sufficient intelligence to perform some task (such as operating a vacuum cleaner), we redefine the task in such a way that the device no longer needs intelligence to accomplish the task (we fuse the vacuum cleaner with the device to such an extent that it can't function as anything but a vacuum cleaner).
(ie, it's a lot harder to build something that can move about to both feed the dog and clean the floor),
Yes. Because that would require intelligence. Intelligence is first and foremost about flexibility and adaptability.
There is no expectation that a computer should work in the same way as a human; that's obviously a rather unfair and biased expectation.
It's the most rational expectation. We can recognize intelligence. We know that humans are intelligent. We know that animals have some degree of intelligence. To postulate that there are other kinds of intelligence as well is a matter of faith. The problem with machine intelligence is that none of the machines appear to us as intelligent.
Also, once we've achieved some area of AI, it's viewed as "obviously trivial", and people point to the remaining things computers can't do, as "proof" that computers aren't intelligent at all.
No, it's just that people are consistently underwhelmed by what AI delivers. The AI community promises a household robot, delivers the Roomba. The AI community promises speech recognition, delivers error-prone transcription software. The AI community promises autonomous cars, delivers a car which completes all of 7 miles in the DARPA grand challenge.
Slashdot Mirror
Yes. This is so true.
What is the purpose of this post? What does it even mean? What is the purpose of posting a link to a nebulous summary of a highly suggestive report on an extremely politically charged subject on a site that bills itself "News for Nerds"?
False dichotomy, based on the unproven assumption that there was anything nefarious going on that required urgent exposition.
You have found Jesus Christ in a potato chip.
With all due respect to an eminent and brilliant physicist, Freeman Dyson is not a climatologist.
But Dyson's argument in fact questions the methods and claims of the science of climatology. If his argument is valid, then it does not matter that he is not a climatologist - it makes Dyson's criticisms more urgent, not less.
All of that is true. But I have come to believe that perhaps the blame lies primarily with the standards themselves. They are just not very good.
I know this is not a popular opinion. Let me qualify it and try to explain briefly what I mean. There is of course a lot of theoretical and historical background to consider, but frankly it is a waste of time to drag all that into a Slashdot comment.
The first problem is with HTML. HTML abstracts at the wrong level. It should be a presentation language, not a structural markup language. There is no need for HTML as a structural markup language and frankly I am baffled by the religious zeal with which some people defend this notion. As a structural markup language, HTML is very poor. Structural markup is most useful for well-defined, domain-specific applications. That is not what HTML is used for and this causes numerous problems: ill-defined rendering behavior, poor querying and indexing abilities, poor feature set, relatively slow performance, not to mention poor reusability.
The second problem is with CSS. Although at its core a good idea, it is poorly implemented, with a pointlessly weird, C-inspired syntax. It is too feeble to express presentational structure and lacks a method to express generalized context-dependent relationships. The selector language is so baroque that it is poorly understood by authors and implementors alike. Most importantly, CSS simply does not solve many common layout and styling problems, except at the most trivial level. Efforts to address this have mostly just made CSS more complicated rather than more powerful.
The third problem is with Javascript. The language itself is not bad, but it exists in an environment that is so primitive and crude, that often the easiest way to accomplish anything is still to just stuff precalculated strings into a node's innerHTML. The web is littered with the corpses of Javascript libraries to provide simple services like data binding, templating, input validation and widget sets. None of which build on eachother, because there are no tools to enable this kind of workflow, and most of which fail on either correctness, performance or standards-compliance.
Why are there so many problems with these standards? Is it normal for standards to be so problematic? It is certainly true that numerous standards failed. But on the other hand, many other standards succeeded. PostScript and PDF are very successful standards that have been implemented dozens of times with minimal interoperability issues. The same goes for countless file format standards, such as GIF, PNG, JPG and ZIP, or standards such as ASCII or Unicode.
Of course the comparison between HTML (and all related tech) and, say, GIF, is not valid. In the case of HTML there are many reasons, some socio-economic, which have brought us to the point where we are today. But despite that, I believe it is possible to identify 2 key issues with the W3C family of tech:
It's probably best if you start off learning what SWFs are and how the Flash "IDE" works before you start looking at the alternatives. It's a pretty complex system with heavy emphasis on graphics and animation and a large amount of cruft. You can trial Flash for 30 days, after that you can look at tools like FDT, MTASC, SWFmill and FlashDevelop. See osflash.org for a variety of open source tools around the Flash platform (like MTASC and SWFmill).
FlashDevelop and the FDT plugin for Eclipse, to name just two, both provide a very good development environment, in many ways better (from a developers perspective) than the Adobe IDE. Of course the Adobe IDE also doubles as a design environment, and in that role it has virtually no competition. But this has traditionally been the case with design tools. Because designers heavily invest in a specific tool (like Photoshop, InDesign or Quark), and those skills are harder to transfer between different tools than the skills of a developer.
Not really. The SWF file format is open, and there are quite a number of quality open-source Flash development tools. There is no open source Flash runtime which is as good as the Adobe runtime, but then there is no Java runtime which is as good as Sun's runtime either.
Most of the commentary on this topic is shit, fueled by ignorance and unthinking dogmatism.
.NET. Many people like to denounce .NET as "Java copied badly" or point out how poorly Windows Forms compares to what's available for GNOME and MacOS X.
.NET are not be as good as other offerings on the market, as a whole there is nothing which compares to it. .NET brings everything under one roof and eliminates entire classes of "glue" and "can't get there from here" problems.
.NET in most cases does The Right Thing. If you haven't worked with .NET yet, just try it and come with specifics. Don't come arguing on abstract principles please.)
.NET 3.0 and WPF, a brand new UI subsystem has been added to the mix, which in terms of raw capability rivals anything out there. ... Christ, that sounds like a commercial. But it's true. You've all seen the demos of movies projected onto flying 3D surfaces etcetera, and this might have left you with the impression that there is little substance to the technology apart from fizz and sparkle.
.NET and WPF form the foundation for the next generation of Windows applications and Silverlight brings parts of this technology to the web. Thus, while Silverlight may falter, as some of you have been suggesting, the underlying technology certainly will not be going anywhere anytime soon.
.NET is that it provides a clean and sane means of unifying traditionally separate realms of development. With .NET and Silverlight, it is slowly becoming possible to leverage the same skills and code on the Web (both server side and client side), the desktop, games consoles, set top boxes, PDAs and Mobile phones.
.NET for some specific task/domain, it would have to be several orders of magnitudes better before the marginal benefit offsets the costs of not being able to ride the slipstream of the Microsoft/.NET juggernaut.
Over the past half decade or so, Microsoft has been developing arguably the most comprehensive and coherent development platform ever on the planet, viz.
This kind of argument is completely besides the point. While some parts of
(Yes, we are all software developers and enthusiasts. We all know the joys of loosely coupled systems and the evils of integration. I'm realy not interested in a generic discussion on that. In practice all good things have costs and all bad things have benefits and
With
That would be a very wrong impression.
Therefore to suggest that Miguel or "we" could or even should be developing an "alternative to Silverlight" is absolute nonsense and indicates an utter blindness for the bigger picture.
The whole point of
Even if you develop something that's significantly better than
Microsoft has been busy rewriting their entire crufty codebase to a modern, unified platform. We are still arguing over widget sets and the relative merits of the GNOME file selector dialog vs. the KDE one. Wake up people.
"Those two headlights heading towards me at high speed could be a bus that will kill me, but they could be two motorcyles that will pass harmlessly on either side of me. I think I'll stand here until I know for sure."
Were it so simple. In your example there is no benefit to remaining where you are and no penalty for stepping aside. In the real world the benefit of the status quo is huge and the penalty for change enormous.
Look, the only reason why Intel and HP keep sinking money into this dog is because they've already spent incredible amounts on it and have been telling everyone, for about a decade now, that it's the Future of Enterprise Computing.
They seem unable or unwilling to face up to the fact that the Itanium is a complete albatross, and there's really nothing admirable about that.
I was referring to Popper's falsification principle. A thousand white swans might support the hypothesis that all swans are white, but just a single black swan disproves it.
Now the problem with the global warming hypothesis is that it has become increasingly resistant to falsification. We are finding a number of black or at least grey swans;
Taken together, these observations raise serious questions about the degree to which we understand global warming and the degree to which we can or should prevent it.
That the Earth is getting warmer on average is probably a given; everyone is familiar with the fact that the temperature in big cities can be up to a full degree or two higher than in the countryside, and combined with the fact that the planet is becoming increasingly urbanized, it stands to reason that the Earth's temperature will have increased by some amount over the last 50 or 100 years.
The real question is whether this is a problem. And to the extent that it is indeed a problem, what is the best solution? It costs billions and billions to effect the few percent CO2 emission reduction as required by Kyoto; but the evidence for CO2 being a causal agent in global warming is very thin; and the ultimate effect of this reduction in terms of e.g. environmental damage or sea level rise is unpredictable and perhaps negligible. Then considering all this, and considering that all the money we sink into emission reductions can no longer be spent on projects that have a directly measurable effect, such as better public transport, or wildlife conservation, or water management; is this really a windmill we want to be tilting at?
There's about 2 million more out there, thanks to a strong consensus by people who do this for a living.
a ta/ushcn/stationoftheweek.jsp, for example.
Again, science is not a consensus discipline. For example, the strong consensus in the '70s was that we were about to enter a new ice age. Nothing of the sort seems to be the case today. In a very real sense, scientific progress comes about not from the 100 corroborating experiments but from the single anomalous one.
The data provided in your link is interesting but the source is obviously biased and every particular claim easily disputed. Your "2 minutes of Googling" could equally have turned up http://www.co2science.org/scripts/CO2ScienceB2C/d
More importantly your link doesn't address the fundamental questions: to what extent is the climate changing, what kinds of effects will this have, to what extent can or should we prevent it?
The original poster, being a fanatic, doesn't entertain any of these questions. For a fanatic, it's only important to know that a) the sky is falling, b) we are to blame, and c) they are morally superior to everyone who doesn't see it that way.
Whereas for normal people, the question is one of compromise: to what extent is climate change actually happening, to what extent can we manage or control the risks of climate change, and to what extent do we sacrifice other desirable goals in order to curb the risk.
I'm sorry. Did you miss the link?
The link, dimwit, contains precisely the sort of babble of which the proximity to reality is being questioned. To wit:
a) Satellite and weatherballoon data doesn't show significant warming
b) Panel insists global warming must nevertheless be occurring
You are asking us to believe that temperature data from two isoloated stations should contradict the overall trend from hundreds of thousands of stations across the entire planet over hundreds of years.
At least he provided data. You provide only babble.
This isn't crazy hippies crying about killing the poor flowers somewhere. This is a near perfect consensus of scientists across dozens of different disciplines on a global scale.
Science is not a consensus discipline. At the same time scientists are not immune to groupthink. The "scientific consensus" has included howlers such as the subhumanity of negros and the notion that every boy wants to fuck his mother.
Just uneducated, dishonest conservatives who see global climate change as a threat to their monied way of life, and who are willing to risk making a world of Katrina-like events commonplace to protect that way of life.
The root of the matter is that you and your cronies fundamentally oppose the wealth and affluence (and, yes, the inherent waste) of the society we live in, and will use any excuse at your disposal to curb its further growth, be it global warming, overpopulation, the dawning of a new ice age, nuclear scares, peak oil or worried aliens.
Don't do this! By using vegetable oils you're depriving the government of tax revenues, so they'll have to raise taxes elsewhere!
(actual argument used by a Dutch mayor)
It's true that planning is hard and software is frequently late, but the allegation here is not that Microsoft is bad at planning. The allegation is that they purposefully and intentionally announce unrealistic schedules in order to string along their customers. By announcing that the next release of Windows is "just around the corner" customers are dissuaded from evaluating the offerings of the competition.
The Vision Quest in Wired 10.09 of September 2002.
The article, as well as the feasibility of Dr. Dobelle's (who has died in 2004) research, are sketchy at best. Apply truckload of salt.
Proposals like these are stupid and idiotic because half the time it simply ISN'T CLEAR what a piece of software should be doing. We simply don't have formally rigorous and perfect descriptions to meet real-world demands, and for the most part we CAN'T have them. If we had, we could use those instead of writing computer programs. For the time being, however, computer programs are the best descriptions we have.
Furthermore, the "intelligence" that a human uses in using a vacuum cleaner is not conscious thought, but comes from "brute force" computation in the brain.
This is a gross misappropriation of terms, because it is highly debatable whether the processes in the brain can be reduced to computational processes. The "brain=hardware, mind=software" metaphor befuddles as much as it explains.
Intelligence should be about end results. If someone else solved a puzzle more quickly than me, it would seem odd to regard him as unintelligent just because he used a different method to me.
This touches on the whole intentionalism debate. Your specific example is flawed because we really do not know what tasks require intelligence. We tend to think intelligence is a prerequisite for solving puzzles because we've only ever seen intelligent beings solve puzzles. But the singular most important discovery made through AI research has been that for many tasks, you don't need intelligence.
If in x years time we have proper speech recognition, but it comes primarily through advancements in hardware rather than more clever algorithms, will speech recognition join chess playing as examples of AI which aren't really intelligent?
If it turns out that you don't need intelligence to recognize speech, sure.
And once this is achieved, people like you will be saying it's not AI, because it doesn't come in the form of a humanoid robot which turns the steering wheel and pushes the pedals!
It's just a driving machine. It seems quite probable that you don't need intelligence to build one. But so far we can't even manage that.
I don't mean to be a sophist. I just think the extentionalist approach to AI, notwithstanding its obvious successes, has more or less exhausted its potential. Because underlying its hard-nosed empiricism is the very magical belief that if only we can create accurate enough simulations, the real thing will spontaneously emerge. Essentially it amounts to cargo cultism.
If it's more efficient to have several dedicated devices rather than an all-in-one, why does this mean AI has failed?
The failure is that instead of building a device with sufficient intelligence to perform some task (such as operating a vacuum cleaner), we redefine the task in such a way that the device no longer needs intelligence to accomplish the task (we fuse the vacuum cleaner with the device to such an extent that it can't function as anything but a vacuum cleaner).
(ie, it's a lot harder to build something that can move about to both feed the dog and clean the floor),
Yes. Because that would require intelligence. Intelligence is first and foremost about flexibility and adaptability.
There is no expectation that a computer should work in the same way as a human; that's obviously a rather unfair and biased expectation.
It's the most rational expectation. We can recognize intelligence. We know that humans are intelligent. We know that animals have some degree of intelligence. To postulate that there are other kinds of intelligence as well is a matter of faith. The problem with machine intelligence is that none of the machines appear to us as intelligent.
Also, once we've achieved some area of AI, it's viewed as "obviously trivial", and people point to the remaining things computers can't do, as "proof" that computers aren't intelligent at all.
No, it's just that people are consistently underwhelmed by what AI delivers. The AI community promises a household robot, delivers the Roomba. The AI community promises speech recognition, delivers error-prone transcription software. The AI community promises autonomous cars, delivers a car which completes all of 7 miles in the DARPA grand challenge.
Timeliness is (or can be) definitely a factor in determining intelligence. We use it as humans all the time.