The pendulum swings back and forth. I wish it would just stay on "social justice,' but it doesn't, because people get complacent
Complacency is one cause. Extremely heavy-handed implementation of "social justice" by some of its loudest advocates (in the last 100 years the Marxist-Lenists, 300 years ago the first French revolutionaries) also did no favours for the leftist cause. The cure was worse than the disease in a lot of cases - to many observers even Hitler seemed like the good guy in comparison to Stalin until WW2.
Don't get me wrong, I'm not particularly fond of capitalism as a system, it's obvious that it's riddled with contradictions, but when it comes to alternatives I'm more of a fan of E F Schumacher than Marx and Lenin. But like organic farming, we've not seen a lot of large-scale implementations of intermediate technology and "Buddhist economics".
Our ideas of social justice are informed by our ideas of what justice is to start with. To the Right, it's obviously just that the more capable and cunning get more. To the Left, it's obviously just that everyone gets an equal share regardless of talent or effort. Both groups often equate justice with violent retribution, and it's just a question of whether morality is privately owned guns shooting those who steal private property, or publically owned guns shooting those who offend the public interest. The question of whether justice is about violence at all is often overlooked.
Has our country been in a situation like this before where all the powers-that-be seem to be working together for their benefit, at the expense of everyone else's freedoms, liberties, and way of life?
Sadly, yes, it's been exactly this way for a while. But previous administrations and corporate heads were much smarter at hiding the fist inside a velvet glove. It's just becoming more nakedly obvious in recent years, as the number of media companies have shrunk, the Internet grassroots has risen, and the outsourced, dematerialised, copyright-based US economy has started seriously wobbling.
The 1980s Reagan years were filled with government and media collusion and outright scandal (Iran-Contra, Reagan saying "Mr Gorbachev, tear down this wall" while Thatcher didn't want it to fall), as were the Cold War 1950s-1960s (COINTELPRO, CIA adventurism in South America), the WW2 1940s were flooded with racist pro-war propaganda ("Smack a Jap", the British Passport Control Office), and the 1920s-30s were drenched in greed and fraud. Then there was WWI, the Spanish-American War, Jim Crow, the Civil War...
But yeah, it feels like the last time things bubbled up to the surface all at once like they're doing now was the 1930s. Not a good feeling at all.
The UN must invade the EU to protect their human rights!
Oh yes. I can see the chilling jackboots striding out now:
"Hi! We've got some blue helmets, a few medical and food aid programs, and a debating chamber in New York which exists entirely at the sufference of the United States. You've got... um... NATO, Russia, Germany, the US's empire of military bases and commercial interests, and nuclear weapons? Righty ho, we'll just drop off these food parcels and be on our way. Sorry to have bothered you! Would you care to make a donation towards our ad campaign to think about someday talking about everyone saying something nice and non-committal about climate change?"
Now, if you meant US/NATO, they've got a solid track record in invading and installing puppet governments sovereign foreign nations.
Having internet access is decidedly not a human right.
What if the Internet becomes a required communication channel in order to be a citizen? What if the local government decides it's cheaper to just post laws on their website than print them?
We consider it normal currently for every child to receive State-funded education in order to read and write their local language. That education, however, is expensive, and it wasn't always considered a "right" for everyone to receive this.
Similarly, blocking someone's access to a bank account is normally reserved only for criminals. We otherwise tend to consider banking services to be a fundamental requirement of being a free citizen.
You could make a case that, if not right now, we are rapidly approaching a world where Internet access is a fundamental prerequisite for engaging in any human social or commercial activity, and not having that access would indeed make a person a social dropout equivalent to being unable to read or write. Hence the term "computer literacy", which equates computer access with education.
You honestly think the taxpayers should foot the entire bill for your YouPorn habit, because you have a RIGHT to YouPorn?
He/she probably does.
I'm guessing GP poster possibly lives in a country (like the USA) where there aren't generally Internet transfer caps, so is assuming that a monthly flat rate pays for the same amount of Internet if they watched hi-def streaming video 24/7, as they logged in once a month to do banking and pay taxes.
A note to Americans: For those of us who live in countries (most of the rest of the world) where transfer caps exist, there's no such thing as simply "getting the Internet" - you get the Internet sold to you in gigabyte chunks, the more you use the more you pay, video is definitely an expensive consumable luxury, and under that charging regime, yes, it seems immoral for some to sit and watch video because others "have to" pay for that habit.
But if Internet were flat-rate monthly, and the cost didn't change depending on what media you watched? The rest of us wouldn't care how much our neighbours used, and would indeed consider Internet access a civil right, since if you didn't get it at all you couldn't talk to the government or your bank.
Some of the most brilliant people I have ever met have severe issues with spelling
Those people are probably not programmers. If they are, then they've got severe work-related problems and they should probably consider a different line of employment.
Seriously, I don't want someone who can't remember function or variable identifiers correctly, coding Internet-facing software. And if you can remember how to spell a variable, you can remember how to spell an English word.
Why do you want to write software, if you don't want to write it correctly?
I'm constantly amused at how many self-proclaimed "programmers" on this website seem to be unable to spell or type the names of textual identifiers consistently. And then complain about "nit-picking" when someone gives them a ?SYNTAX ERROR.
How do you manage to get your programs to compile? Just keep randomly retyping keywords until the build script accidentally succeeds?
Ah, youth of today. I remember typing in BASIC listings from magazines where a one-character mistype in 100 lines of random hex DATA statements would mean you couldn't play Attack of the Mutant Space Zombies. And that would be terrible. We learned to type, darnit, and we learned to proofread, and we sure as heck learned to spell.
Additive manufacturing, or accretion printing, isn't wasteful. But having the ability to recycle printed parts back into raw plastic would be the big issue in space.
Not everybody believes in "spend money to make money".
Or even "first do no harm", which in the software world translates to "first install no botnets on your customers' machines". Certainly Microsoft, Apple, Adobe and Sun/Oracle don't, since they're the top placers for drive-by botnet installs. But none of the open source projects are immune either - Mozilla and Linux still push out monthly security updates.
What worries me most about the present (let alone the future) of software quality testing is that we still don't even have a way to tell if we're doing harm when writing code. Merely proving program correctness in the very restricted realm of "no random filesystem or TCP/IP packet data executed as machine code" seems to be beyond us. Seriously how hard can that be? How many programs, other than just-in-time compilers, need the ability to literally transform user data into machine code AND then run it in the same process context? Why do these kinds of exploits keep occurring in purely data-transform software like PDF or JPEG parsers? Why do our current best-of-breed programming languages seemingly make it impossible to simply assert "this program will NEVER need to dynamically execute as raw machine code ANY incoming data?" Why do our current best-of-breed operating systems allow programs to communicate through raw machine code in the first place? Wasn't the entire point of the object-oriented revolution, from Smalltalk on, that we would build systems out of message-passing and "shared-nothing" designs? Yet we've apparently abandoned the fundamental principle of OOP while still keeping the abstract formality.
I remember when Java was first announced, circa 1995, and its main selling point was "absolute security" since it was a virtual machine. I'm still scratching my head as to how they managed to cram a type-safe, pointer-free VM so full of critical vulnerabilities. I mean, that should be a theoretical impossibility, shouldn't it?
I'll believe that when I stop getting monthly Critical security updates from Microsoft, Apple, Oracle and Adobe for products which already passed their entire battery of industry-standard best-practice software QA tests.
Philosophically speaking, I'm not actually sure that it's possible to deprive anything of its selfhood, short of retroactively erasing it from time and space (or total Buddhist enlightenment, perhaps).
But philosophical self-existence is not actually a very useful indicator of whether you've done any harm to a person or animal.
Websites have implemented custom, animation-heavy interfaces for decades via Flash and now HTML5.
And despite that marketeers and web designers keep trying to hawk them as improvements, those elements of those websites are universally awful and reviled.
Remember what made Google successful. A simple, clean design in the GeoCities era of flashing background "under construction" GIFs.
You young'uns maybe don't remember the 90s, let alone the 80s, but trust me - things have been a lot worse in the "every app does it differently" direction in which we're currently trending back towards. As users we really don't want to repeat those past UI mistakes, but the current crop of designers seem hell-bent on it.
Oh well. Another twenty years and the clean and classic look should be back in fashion.
Is a numerical simulation of a thing, the same as an actual instance of a thing ?
Does a submarine swim?
Not at all in the same sense a fish does, if by "swim" you mean "display the full range of behaviours exhibited by the lifeform one is claiming to simulate". It doesn't mate, grow, eat or self-repair. (You could perhaps make the case that the submarine-crew gestalt, seen as a cybernetic organism, does display some of these behaviours, but you said "submarine" which generally means just the physical shell.)
A submarine does, however, displace and move through water like a rock does. Well done. You've successfully simulated one tiny aspect of a lifeform's vast repertoire by reducing it to a triviality: physical motion. By using the same technique I''ve also successfully demonstrated that a submarine and a rock are identical.
However, as a philosophical point about the equivalence of computing and thought, this answer comes a little short.
Is a numerical simulation of a thing, the same as an actual instance of a thing ?
Is a numerical simulation of cutting 3D shapes on a CAD/CAM machine the same as actually cutting 3D shapes on a CAD/CAM machine?
Well, it is if you plug the output of the simulation into the machine. It's not if you don't. In other words, _actions_ that can be controlled by a computer, are as real as any other action. Simulations of those actions aren't.
The question you should be asking is, _can_ a computer create actions which are as complex (and therefore as real) as the behaviours displayed by lifeforms?
Currently, no. There are two limits to what lifelike behaviours a computer can produce: one in hardware, one in software.
The hardware limitation is that a computer has no true analogue of a DNA based lifeform's cellular structure. Computers and industrial robots are built from very large components, and are very far away from being able to synthesise themselves at the microscopic level out of readily available components like lifeforms can. Possibly nanotechnology might give computers this ability; I'm skeptical. Even building something like Terminator's Skynet - a computer capable of building self-maintaining robots out of automated factories - would actually be very hard. It would need a complete global supply chain including mines, smelters, refineries, and research laboratories before it could come close to what a living cell does every second. And without a cellular structure no computer is going to truly reproduce some of the fundamental features of life: conception, growth, feeding, self-repair.
(There's a possible workaround for this limitation. One of the big ideas introduced by cyberpunk fiction in the 1980s was the "matrix" - that a sufficiently complex shared virtual reality might allow software-only robots to ignore the "hardware gap" and still experience the essential features of life. This is still a strong meme in today's postcyberpunk SF. Is it feasible? Maybe. How real could World of Warcraft get, and at what point would a sufficiently self-aware bot be able to decide that this wasn't a "real" reality it inhabited?)
The second problem is probably harder: the software problem. Currently, we simply don't have any idea how to build a truly self-aware AI, and that's interesting in itself. Our best efforts in both computing and cognitive psychology show our huge lack of data about how our own minds work. We've got models and speculations, but nothing actually implementable, and the successful "AI" folks are mostly just doing big searches and not worrying at all about how realistic a simulation of the human mind they're doing.
In other words, at the moment we can't even _simulate_ a mind, let alone get to the point where we'd have to ask whether the simulation was real or not. And there's good reason to expect that we might never be able to simulate a human mind, that the problem will simply grow in scale and complexity the more computing power we throw at it (since that's been our experience in the past). We can simulate certain small problems, like chess and language translation; what we can't do yet is put them in the sort of personal and emotional context which a human mind brings. And I suspect we never will, that the world of dreams and emotions and other squishy human drives will turn out to be vastly bigger - more complex and more real, in fact - than the world of logical games at which computers excel.
And, if we reach the point where this becomes obvious, that would be a very interesting result. What is it which makes the human mind so big that it saturates our brute-force computer processing power? And why is it that the "bigness" of the human mind seems to reside more clearly in the _ordinary_ things we do - things which anyone can do, like dreaming or language acquisition before the age of 3 - rather than in what we generally consider expensive, specialist expert knowledge, like mathematics or wargaming or chess?
No, not really. A computer is, at the fundamental level, no more complicated than a player piano. Anything we generally refer to as information, has been in fact, programmed ( added / input ) into the computer by a human. It can store the result values of a computation, but in order to do it, the results of the computation had to be input in the compact form which is a program.
That's not strictly true - computers have I/O ports through which they can observe their environment and therefore add new information to their memory which was not part of the initial program load. That information doesn't have to come from a human at all. That entire field is called Machine Learning and is a critical part of today's AI, including Google.
However Machine Learning has never become the general-purpose problem solving paradigm the early AI boosters confidently expected it to be. In the 1950s-60s there was a lot of optimism that a simple universal learning algorithm would be discovered which would allow computers to be fully self-programming (while also being useful). See for instance: General Problem Solver, STELLA (and PURR-PUSS) and the Perceptron (which later was resurrected as Neural Networks). There seems to be no a priori reason why a general learning machine shouldn't exist; the reality however is that we don't tend to find fully self-programming systems useful. AI today tends to make limited use of Machine Learning strategies only as components within a much more rigid, "programmed" framework.
It's interesting though to ask what would happen if we did succeed in building a completely closed box robot which had just a tiny self-programming firmware and a big RAM, and deduced everything else from observing and interacting with its environment. It would be fair to say that it would likely act very different from the programmed computers we know today; it would not necessarily have anything corresponding to "subroutines", "control structures", "databases". Its "code" would be a big messy spaghetti of mixed data and inferences and would not be maintainable by a human. It might not even be able to logically justify its actions or reason about its ideas and might act out what an observer might consider to be "instincts" or "emotions". Its high level structure would not be relational, object-oriented, or match any other programming paradigm we humans use. But at a low level, it could still be a von Neumann machine on an i386 platform.
Slashdot Mirror
This "Federal Fraud" program for launching commercial payloads into LEO seems legit. Tell me more!
The Ballistic Missile Submarines look cheaper though. How much do they charge per launch?
you'd see someone in the office who will be even more communist than Obama
Zombie Dwight Eisenhower perhaps? He was pretty hardcore commie, by 2011 standards.
The pendulum swings back and forth. I wish it would just stay on "social justice,' but it doesn't, because people get complacent
Complacency is one cause. Extremely heavy-handed implementation of "social justice" by some of its loudest advocates (in the last 100 years the Marxist-Lenists, 300 years ago the first French revolutionaries) also did no favours for the leftist cause. The cure was worse than the disease in a lot of cases - to many observers even Hitler seemed like the good guy in comparison to Stalin until WW2.
Don't get me wrong, I'm not particularly fond of capitalism as a system, it's obvious that it's riddled with contradictions, but when it comes to alternatives I'm more of a fan of E F Schumacher than Marx and Lenin. But like organic farming, we've not seen a lot of large-scale implementations of intermediate technology and "Buddhist economics".
Our ideas of social justice are informed by our ideas of what justice is to start with. To the Right, it's obviously just that the more capable and cunning get more. To the Left, it's obviously just that everyone gets an equal share regardless of talent or effort. Both groups often equate justice with violent retribution, and it's just a question of whether morality is privately owned guns shooting those who steal private property, or publically owned guns shooting those who offend the public interest. The question of whether justice is about violence at all is often overlooked.
Man. you know your history, it is a rare thing, only if the rest of (300 millions -1) Americans were just like you....just imagine......
Not that I want to reinforce a stereotype, but I'm not an American. :)
Has our country been in a situation like this before where all the powers-that-be seem to be working together for their benefit, at the expense of everyone else's freedoms, liberties, and way of life?
Sadly, yes, it's been exactly this way for a while. But previous administrations and corporate heads were much smarter at hiding the fist inside a velvet glove. It's just becoming more nakedly obvious in recent years, as the number of media companies have shrunk, the Internet grassroots has risen, and the outsourced, dematerialised, copyright-based US economy has started seriously wobbling.
The 1980s Reagan years were filled with government and media collusion and outright scandal (Iran-Contra, Reagan saying "Mr Gorbachev, tear down this wall" while Thatcher didn't want it to fall), as were the Cold War 1950s-1960s (COINTELPRO, CIA adventurism in South America), the WW2 1940s were flooded with racist pro-war propaganda ("Smack a Jap", the British Passport Control Office), and the 1920s-30s were drenched in greed and fraud. Then there was WWI, the Spanish-American War, Jim Crow, the Civil War...
But yeah, it feels like the last time things bubbled up to the surface all at once like they're doing now was the 1930s. Not a good feeling at all.
> GO POUND SAND
The pound sand doesn't appear to have an entrance.
The UN must invade the EU to protect their human rights!
Oh yes. I can see the chilling jackboots striding out now:
"Hi! We've got some blue helmets, a few medical and food aid programs, and a debating chamber in New York which exists entirely at the sufference of the United States. You've got... um... NATO, Russia, Germany, the US's empire of military bases and commercial interests, and nuclear weapons? Righty ho, we'll just drop off these food parcels and be on our way. Sorry to have bothered you! Would you care to make a donation towards our ad campaign to think about someday talking about everyone saying something nice and non-committal about climate change?"
Now, if you meant US/NATO, they've got a solid track record in invading and installing puppet governments sovereign foreign nations.
Having internet access is decidedly not a human right.
What if the Internet becomes a required communication channel in order to be a citizen? What if the local government decides it's cheaper to just post laws on their website than print them?
We consider it normal currently for every child to receive State-funded education in order to read and write their local language. That education, however, is expensive, and it wasn't always considered a "right" for everyone to receive this.
Similarly, blocking someone's access to a bank account is normally reserved only for criminals. We otherwise tend to consider banking services to be a fundamental requirement of being a free citizen.
You could make a case that, if not right now, we are rapidly approaching a world where Internet access is a fundamental prerequisite for engaging in any human social or commercial activity, and not having that access would indeed make a person a social dropout equivalent to being unable to read or write. Hence the term "computer literacy", which equates computer access with education.
You honestly think the taxpayers should foot the entire bill for your YouPorn habit, because you have a RIGHT to YouPorn?
He/she probably does.
I'm guessing GP poster possibly lives in a country (like the USA) where there aren't generally Internet transfer caps, so is assuming that a monthly flat rate pays for the same amount of Internet if they watched hi-def streaming video 24/7, as they logged in once a month to do banking and pay taxes.
A note to Americans: For those of us who live in countries (most of the rest of the world) where transfer caps exist, there's no such thing as simply "getting the Internet" - you get the Internet sold to you in gigabyte chunks, the more you use the more you pay, video is definitely an expensive consumable luxury, and under that charging regime, yes, it seems immoral for some to sit and watch video because others "have to" pay for that habit.
But if Internet were flat-rate monthly, and the cost didn't change depending on what media you watched? The rest of us wouldn't care how much our neighbours used, and would indeed consider Internet access a civil right, since if you didn't get it at all you couldn't talk to the government or your bank.
Some of the most brilliant people I have ever met have severe issues with spelling
Those people are probably not programmers. If they are, then they've got severe work-related problems and they should probably consider a different line of employment.
Seriously, I don't want someone who can't remember function or variable identifiers correctly, coding Internet-facing software. And if you can remember how to spell a variable, you can remember how to spell an English word.
Why do you want to write software, if you don't want to write it correctly?
Once again nit-picking spelling
I'm constantly amused at how many self-proclaimed "programmers" on this website seem to be unable to spell or type the names of textual identifiers consistently. And then complain about "nit-picking" when someone gives them a ?SYNTAX ERROR.
How do you manage to get your programs to compile? Just keep randomly retyping keywords until the build script accidentally succeeds?
Ah, youth of today. I remember typing in BASIC listings from magazines where a one-character mistype in 100 lines of random hex DATA statements would mean you couldn't play Attack of the Mutant Space Zombies. And that would be terrible. We learned to type, darnit, and we learned to proofread, and we sure as heck learned to spell.
I blame the iPad myself.
I think you mean secede, not "succeed"
Nothing secedes like secession.
Additive manufacturing, or accretion printing, isn't wasteful. But having the ability to recycle printed parts back into raw plastic would be the big issue in space.
There's a simple solution to this problem:
Space Lego.
Not everybody believes in "spend money to make money".
Or even "first do no harm", which in the software world translates to "first install no botnets on your customers' machines". Certainly Microsoft, Apple, Adobe and Sun/Oracle don't, since they're the top placers for drive-by botnet installs. But none of the open source projects are immune either - Mozilla and Linux still push out monthly security updates.
What worries me most about the present (let alone the future) of software quality testing is that we still don't even have a way to tell if we're doing harm when writing code. Merely proving program correctness in the very restricted realm of "no random filesystem or TCP/IP packet data executed as machine code" seems to be beyond us. Seriously how hard can that be? How many programs, other than just-in-time compilers, need the ability to literally transform user data into machine code AND then run it in the same process context? Why do these kinds of exploits keep occurring in purely data-transform software like PDF or JPEG parsers? Why do our current best-of-breed programming languages seemingly make it impossible to simply assert "this program will NEVER need to dynamically execute as raw machine code ANY incoming data?" Why do our current best-of-breed operating systems allow programs to communicate through raw machine code in the first place? Wasn't the entire point of the object-oriented revolution, from Smalltalk on, that we would build systems out of message-passing and "shared-nothing" designs? Yet we've apparently abandoned the fundamental principle of OOP while still keeping the abstract formality.
I remember when Java was first announced, circa 1995, and its main selling point was "absolute security" since it was a virtual machine. I'm still scratching my head as to how they managed to cram a type-safe, pointer-free VM so full of critical vulnerabilities. I mean, that should be a theoretical impossibility, shouldn't it?
China is the Han race.
The Han shot first!
I'll believe that when I stop getting monthly Critical security updates from Microsoft, Apple, Oracle and Adobe for products which already passed their entire battery of industry-standard best-practice software QA tests.
no one else has a right to deprive me of my self
Philosophically speaking, I'm not actually sure that it's possible to deprive anything of its selfhood, short of retroactively erasing it from time and space (or total Buddhist enlightenment, perhaps).
But philosophical self-existence is not actually a very useful indicator of whether you've done any harm to a person or animal.
that damned Philadelphia Experiment. Will mankind ever learn?
Einstein tried to warn us that cream cheese warps space and time, but we went right ahead and put it on sandwiches.
Now Hitler's got cheesesteak technology. I hope you scientists are happy.
What is frelling wrong with you people?! What kind of lame expletive is "frack" ?
Smee. Smeeeeeeeeee. Smeeeeeeeeeeeee-
Excuse me, I'm dreadfully sorry for that outburst. POLITENESS PROTOCOLS DISABLED.
Smeghead.
And we know where you were last Summer...
But you don't know where I'll be next.
And my name's not Summer.
Websites have implemented custom, animation-heavy interfaces for decades via Flash and now HTML5.
And despite that marketeers and web designers keep trying to hawk them as improvements, those elements of those websites are universally awful and reviled.
Remember what made Google successful. A simple, clean design in the GeoCities era of flashing background "under construction" GIFs.
You young'uns maybe don't remember the 90s, let alone the 80s, but trust me - things have been a lot worse in the "every app does it differently" direction in which we're currently trending back towards. As users we really don't want to repeat those past UI mistakes, but the current crop of designers seem hell-bent on it.
Oh well. Another twenty years and the clean and classic look should be back in fashion.
Does a submarine swim?
Not at all in the same sense a fish does, if by "swim" you mean "display the full range of behaviours exhibited by the lifeform one is claiming to simulate". It doesn't mate, grow, eat or self-repair. (You could perhaps make the case that the submarine-crew gestalt, seen as a cybernetic organism, does display some of these behaviours, but you said "submarine" which generally means just the physical shell.)
A submarine does, however, displace and move through water like a rock does. Well done. You've successfully simulated one tiny aspect of a lifeform's vast repertoire by reducing it to a triviality: physical motion. By using the same technique I''ve also successfully demonstrated that a submarine and a rock are identical.
However, as a philosophical point about the equivalence of computing and thought, this answer comes a little short.
Is a numerical simulation of a thing, the same as an actual instance of a thing ?
Is a numerical simulation of cutting 3D shapes on a CAD/CAM machine the same as actually cutting 3D shapes on a CAD/CAM machine?
Well, it is if you plug the output of the simulation into the machine. It's not if you don't. In other words, _actions_ that can be controlled by a computer, are as real as any other action. Simulations of those actions aren't.
The question you should be asking is, _can_ a computer create actions which are as complex (and therefore as real) as the behaviours displayed by lifeforms?
Currently, no. There are two limits to what lifelike behaviours a computer can produce: one in hardware, one in software.
The hardware limitation is that a computer has no true analogue of a DNA based lifeform's cellular structure. Computers and industrial robots are built from very large components, and are very far away from being able to synthesise themselves at the microscopic level out of readily available components like lifeforms can. Possibly nanotechnology might give computers this ability; I'm skeptical. Even building something like Terminator's Skynet - a computer capable of building self-maintaining robots out of automated factories - would actually be very hard. It would need a complete global supply chain including mines, smelters, refineries, and research laboratories before it could come close to what a living cell does every second. And without a cellular structure no computer is going to truly reproduce some of the fundamental features of life: conception, growth, feeding, self-repair.
(There's a possible workaround for this limitation. One of the big ideas introduced by cyberpunk fiction in the 1980s was the "matrix" - that a sufficiently complex shared virtual reality might allow software-only robots to ignore the "hardware gap" and still experience the essential features of life. This is still a strong meme in today's postcyberpunk SF. Is it feasible? Maybe. How real could World of Warcraft get, and at what point would a sufficiently self-aware bot be able to decide that this wasn't a "real" reality it inhabited?)
The second problem is probably harder: the software problem. Currently, we simply don't have any idea how to build a truly self-aware AI, and that's interesting in itself. Our best efforts in both computing and cognitive psychology show our huge lack of data about how our own minds work. We've got models and speculations, but nothing actually implementable, and the successful "AI" folks are mostly just doing big searches and not worrying at all about how realistic a simulation of the human mind they're doing.
In other words, at the moment we can't even _simulate_ a mind, let alone get to the point where we'd have to ask whether the simulation was real or not. And there's good reason to expect that we might never be able to simulate a human mind, that the problem will simply grow in scale and complexity the more computing power we throw at it (since that's been our experience in the past). We can simulate certain small problems, like chess and language translation; what we can't do yet is put them in the sort of personal and emotional context which a human mind brings. And I suspect we never will, that the world of dreams and emotions and other squishy human drives will turn out to be vastly bigger - more complex and more real, in fact - than the world of logical games at which computers excel.
And, if we reach the point where this becomes obvious, that would be a very interesting result. What is it which makes the human mind so big that it saturates our brute-force computer processing power? And why is it that the "bigness" of the human mind seems to reside more clearly in the _ordinary_ things we do - things which anyone can do, like dreaming or language acquisition before the age of 3 - rather than in what we generally consider expensive, specialist expert knowledge, like mathematics or wargaming or chess?
Living thing: a 3D pattern that attempts to reproduce itself.
So, fire is a life form then.
No, not really. A computer is, at the fundamental level, no more complicated than a player piano. Anything we generally refer to as information, has been in fact, programmed ( added / input ) into the computer by a human. It can store the result values of a computation, but in order to do it, the results of the computation had to be input in the compact form which is a program.
That's not strictly true - computers have I/O ports through which they can observe their environment and therefore add new information to their memory which was not part of the initial program load. That information doesn't have to come from a human at all. That entire field is called Machine Learning and is a critical part of today's AI, including Google.
However Machine Learning has never become the general-purpose problem solving paradigm the early AI boosters confidently expected it to be. In the 1950s-60s there was a lot of optimism that a simple universal learning algorithm would be discovered which would allow computers to be fully self-programming (while also being useful). See for instance: General Problem Solver, STELLA (and PURR-PUSS) and the Perceptron (which later was resurrected as Neural Networks). There seems to be no a priori reason why a general learning machine shouldn't exist; the reality however is that we don't tend to find fully self-programming systems useful. AI today tends to make limited use of Machine Learning strategies only as components within a much more rigid, "programmed" framework.
It's interesting though to ask what would happen if we did succeed in building a completely closed box robot which had just a tiny self-programming firmware and a big RAM, and deduced everything else from observing and interacting with its environment. It would be fair to say that it would likely act very different from the programmed computers we know today; it would not necessarily have anything corresponding to "subroutines", "control structures", "databases". Its "code" would be a big messy spaghetti of mixed data and inferences and would not be maintainable by a human. It might not even be able to logically justify its actions or reason about its ideas and might act out what an observer might consider to be "instincts" or "emotions". Its high level structure would not be relational, object-oriented, or match any other programming paradigm we humans use. But at a low level, it could still be a von Neumann machine on an i386 platform.