With regards to NVidia - well, what they want in an employee is someone who's not stupid and who is willing to do some work. A lazy genius is no use. That sort of person will get a good GPA. As regards the test thing - well, if you mean they insist you do it and blacklist you if you don't, then that is being a prick. But if they're simply allowing you to do a first screening (and a lot of students will go to a career fair expecting to be there all day and do have time) on the spot rather than waste money going to their HQ half way across the country, I don't see any problem.
For what it's worth, IMO if you interview someone who you'd be happy to work with but lacks technical skills, you probably should tell them - it doesn't take long to give them an idea of where their problems are. If you have someone you think is a prick, then whether or not their technical skills are up to scratch just give them the form letter.
Notes remaining legal tender isn't really all that uncommon. What everyone does is to simply stop reissuing the old variety, so every time an old style note goes into a bank it gets destroyed, and a new style note comes out in its place. Give it six months and you still see a few old style notes about, but not very many - think about how long you normally hang onto a note for: a few days at most. This is also why you don't see too many ripped and tattered notes about, banks destroy them when they get them.
No, of course they can't. The idea is that a lot of developers who don't want to write an Office competitor anyway may well be interested in the freebie: it makes their product look like an MS product and it's (presumably) easier to use the freebie than write your own MS-style UI code. For the average customer, MS-style means it looks professional. Which is a big selling point.
What MS get out of this is that when you have a desktop full of applications which all have one UI style, any other style looks out of place, so the competition looks bad. Same deal as with VB: you can write your app in VB and it does a lot of GUI creation for you, but what you get will be MS-style buttons and so on.
No. OK, so the way fusion is known to work is this: you have nuclei which are moving fast, they smash together and fuse. This gives off energy.
Now, the way this guy's done it (which is not original in method, but technically the guy's got to know what he's doing) is that you use an electric field to accelerate the nuclei. The problem with this is that you can't (easily, anyway) either accelerate large numbers of nuclei or keep the (vast) number which don't fuse from losing most of their energy to the environment. The result is that although the total energy out is larger than the total in, it's only by a very small fraction, and if you try to generate electricity from the heat you end up with less than you used to accelerate the nuclei. Generators aren't 100% efficient (per thermodynamics, they can't be even in theory).
In the Sun, you have a whole lot of high energy nuclei (it's hot near the centre of the Sun), so that when a couple of especially energetic nuclei collide they fuse; when they miss energy isn't dissipated much because they simply bounce off other energetic nuclei. The result is that fusion keeps going without energy needing to be put in all the time. This is where we have a technical problem now: we can't keep fusion going without putting in vast amounts of energy, and until we can reduce the energy input (to at least a sensible percentage of the total energy output, like 50% instead of 99.9%) we won't have any fusion power stations.
Cold fusion is different: the idea is that you can persuade nuclei to fuse when they don't have enough energy for the conventional method. This isn't impossible in theory, but in practice it hasn't been made to work even with lots of energy input.
The other problem with mail votes is the possibility of vote selling. Particularly if you allow this override-later thing.
Suppose you and your wife vote by mail, and then the final count has the people you voted for win by one vote.
You could probably get quite a bit of money by offering to override your votes, on the basis that an unidentifiable masked man made you do it at gunpoint (how can they prove you're lying?).
Override-later still doesn't get rid of the possibility of intimidation, anyway. OK, you can come along later and say someone made you vote this way - but either you let people do that without following up on the accusation, in which case a lot of people will do it because they changed their minds, and potentially some people will do it because they were told to, or alternatively you follow up on accusations, in which case the police probably get swamped with malicious accusations (my vote isn't worth much, and I can use it to get my hated neighbour...) and in the real cases of intimidation no-one changes their vote anyway on pain of broken arms. Same deal with people giving evidence against the Mafia - even with police protection witnesses were kept quite fairly effectively.
The only way you can really avoid intimidation is to have the vote be a one shot, in secret thing, so no-one but you can find out after the fact how you voted. That being the case, since we can now do both an electronic and a paper vote and compare results, it's probably a good idea to do both since it is then much harder to rig the system (and you can announce the electronic count straight away and it isn't much more work than a standard paper vote).
Except that it's very easy to intimidate with vote by mail. You just find someone who is, go round to their house with a gun and make them fill it in as you want it.
Paper ballots have always had problems with ballot box stuffing and so on. Electronic ballots are easily tampered with electronically (and despite what some people on this site seem to think, you cannot build in voter verifiability without losing anonymity).
If you wanted to make it hard to rig an election, IMO the best way to do this is to have an electronic count and a paper count. Use voting machines in polling stations which send the vote to a central computer electronically and print off a paper copy which the voter picks up, verifies and puts in a ballot box themselves. Then you get the results quickly from the electronic system, and the next day you go and count all the paper votes (with scrutineers from all parties) and check that the electronic count and paper count at each polling station match. That way, to fiddle the count you'd have to tamper with the electronic system and with the paper system, and you'd have to get everything to match up. It wouldn't even be much more work than the standard paper votes system.
Not really. You can issue a crypto receipt of vote, fine. Let's assume that the crypto system is unbreakable, since it doesn't really matter. However the voter's arm is not unbreakable, and you have to ensure that someone isn't standing behind the voter either forcing them to vote this way, or after the election forcing them to check their vote and show it's as required.
Even if you can avoid that problem, then you have the problem that you have a central counting machine, which has an encrypted record of the votes. But it cannot add these up unless it can break the crypto - and that would mean that everyone else could, too. So the individual voter can verify that the central machine received his vote correctly, but he cannot verify that it counted the vote correctly. For example, it wouldn't be too hard to write a program which changed the vote counting code so that every 100th Democrat vote was counted as Republican all day, changed it back to the original after voting closed then wiped itself. And you couldn't then discover that without getting everyone to check their votes and adding up the results - and that would mean everyone would know who voted for what.
Godel's theorems have nothing to do with this. You obviously don't understand the meaning of either 'complete' or 'consistent' in the mathematical logic sense - which is NOT the same as the common English sense.
Godel's theorems do not say 'you cannot do this with computers', they say 'you cannot do this'. Whether you have a hunk of dead tree or a hunk of silicon involved is totally irrelevant.
A paper ballot is obviously proof against electronic tampering. That doesn't make it perfect - after all, if you keep your money stuffed in a mattress it's safer from hackers than if you had it in a bank (as electronic records). But in the former case you will lose everything if a thief comes round.
The problem you have is that you want an election system where only the voter knows how they voted (to avoid intimidation et cetera) but also the counting is verifiably accurate. Unfortunately these aren't compatible. Somewhere there has to be a compromise.
Traditional paper ballots have the problem that you can't detect intelligent tampering with a small number of votes (which can be enough to swing a tight election). Ballot box stuffing, et cetera, are not new.
An intelligently designed electronic system would be reasonably good - i.e. you could easily set up a paper trail which would allow the voter to check that the machine had correctly registered their vote, with not too much trouble you could set up a cryptographic check to ensure that the vote correctly reached a central counting computer. But you couldn't verify that the computer had then performed its job correctly without losing the voter anonymity. Open source counting software would make it slightly harder to fiddle, but it would not be very hard to write a program which runs, fiddles the count then deletes itself afterwards.
The problem with the current US system is that it is not well designed (or at least it's designed by people whose primary concern is profits not a secure system, and they're doing well at that).
There is a difference between 'I want the world to be like this' and 'The world is like this'.
The last time the monarchy in the UK had any power, there was no such thing as the USA. If you want to continue to argue this, you can produce an example of the monarch using their supposed powers any time in the last century or so first, please.
'Inalienable rights' is still just a phrase. Paine's variant (and there have been several) is simply the viewpoint of liberal English society around his time. If what you mean is 'Paine's definition doesn't change over time' then please say so. I'll point out that there are a whole lot of things we'd probably both feel we have a right to which Paine's definitions don't cover. Such as you'd probably object to the police if I set up a giant speaker shouting abuse at you just outside your house. The fact that the police would undoubtedly haul me off for causing a breach of the peace were I to do so is a reflection of the fact that society as a whole does not like people being obnoxious.
As far a 'nonhumans cannot think' - well, firstly I see no barrier to recreating a human brain structure by computer simulation (it'd be prohibitively expensive and probably not practically useful but not impossible) and such a thing would certainly pass the Turing test, assuming you didn't pick someone really stupid to copy the brain structure from - secondly, in terms of society, if you prefer you can read 'majority preference of individuals in society' for 'think'. It doesn't affect my argument.
Gordon Brown, who said the thing you originally objected to, is the Chancellor of the Exchequer of the UK. That is not part of the USA. So the US constitution is totally irrelevant.
As regards elections - the details are different, the result is similar. You may have voted as a party member for your political party to put forward a Presidential candidate, but you have never voted directly for a Presidential candidate. You voted for a member of the electoral college who you knew would vote for the Presidential candidate of your choice (and he has the right to change his mind and vote for the other guy, but this almost never happens). In the UK, I vote for an MP who has duties above and beyond getting the guy I want to be prime minister in, but generally people do not vote for an MP because of who he is but for the MP because of the party leader he represents. Put it another way, when did you last have a President who was not the party political choice of one of your two major parties?
Yes, British people are officially subjects of the Queen. Which means nothing in reality. The Queen has no real power, she simply does as told to by the elected government.
As regards 'we made our government protect our rights' - yes, I would really love to be living in a country where the Patriot Acts are well on the way to creating an outright police state, et cetera. Wake up and look at what your goverment is actually doing.
'Inalienable rights of humanity' - this simply translates as 'we can do what society in general thinks is OK'. It sounds good, but your 'inalienable rights' change over time. Pre civil war, every man had the inalienable right to liberty, except if they were black. Post civil war, things were different. If your society decides it doesn't like your inalienable right to religious freedom (many countries around the world) then talking about your 'inalienable rights' will simply make you look pseudo-intellectual as you're dragged to the gallows. Paine wasn't saying 'this is the way things are', he was saying 'this is how things should be'. There is a very big difference.
You know how the UK is not actually part of the USA? That means we do not follow your laws or constitution.
If you want to go on about the UK, please try to understand that.
Furthermore, we have a thing called elections, that being where the guy with the most votes is the prime minister (like the US, really). Except our version of elections doesn't involve whoever is elected receiving huge 'campaign contributions', AKA bribes, from big companies. Which means we are not so badly affected as you are by political lobbyists. Gordon Brown can actually come out and say things like 'the record companies are morally wrong here' and not risk losing any campaign funds for the next election.
If you want to talk about rights - we elect the government, then they pass laws. If we don't like the laws, we elect a government that will repeal those laws. We, the people, decide what rights we think we should have. That is how goverment works: you are not allowed to do something if your fellow people disagree. The alternative is called anarchy, and it does not work in a civilised society (more accurately, if the government were to vanish you would quickly get rule by the strongest, and that eventually would become a democracy in a country with sufficient natural resources).
There's a big difference between 'funded by license fees which the government requires everyone to pay' and 'government controlled'. The BBC is not known for being friendly to the government. If anything, it goes out of its way to demonstrate its independence, it tends to be more attacking than ITV.
VOA _is_ government controlled, it's essentially a news service that's grown up out of a news/propaganda service that was broadcast mainly for the benefit of the Russians.
In any case, if you actually did something like that to a Diebold machine in the US, Diebold would descend upon you with the full force of the DMCA, try to shut you up, and not do a thing about the machines.
Then the Republican Party would happily go on to violate said DMCA by using your vote-rigging code without permission.
"Even complicated proofs involve an extraordinary amount of simple symbol-pushing that a computer could do easily."
Some do, some don't. Yes, there are occasionally (boring) papers publised where someone has simply done a load of algebra to get a result. But most serious mathematics does not work that way. There are a few people trying to formalise proofs so that a computer can at lest verify them, and that hasn't really got very far (a computer can verify simple logic, but a 2 page proof written out in simple logic could fill a book, so no-one does that). Computers can be useful, but mainly for checking lots of simple examples.
You can easily program a computer to perform algebraic manipulations (MAPLE et al.). This is not the same as producing serious proofs. I doubt a computer will be able to replace a professional mathematician any sooner than it will replace a (good) professional translator, or write a good book.
Our brains may well be (probably are) not deterministic, at least in so far as there will be a whole lot of atomic-level effects that most likely do affect our thought processes. That doesn't mean there is no way for a computer to emulate it. There is no reason why you cannot connect a Geiger counter up to a computer and use that to give you any random bits you might want.
The basic problem is that AI researchers go down one of two routes.
One, they try to get something that looks like intelligence, or is actually useful (game AI, say). Which means they do something which is far too rigid for 'genuine AI', whatever that's meant to be. Essentially, a big list of rules plus some kind of selection system involving a bit of (pseudo-)randomness and a bit of scoring by past results, which they call learning.
Two, they try to set up something which might have enough flexibility to include the possibility of genuine thought. They don't know how to teach it (humans do not magically acquire abilities, we learn by example and reward, and how do you simulate that?), they don't know how to set up a structure which can learn (why is one brain better at learning than another? why do we have the ability to learn a complex language when a monkey does not?). In a couple of years their grant runs out and they have nothing resembling a useful result, so no new grant, game over.
The first route will never get anywhere with producing 'intelligence'. It produces stuff like good chess programs, conversation boxes, et cetera; all of these have improved over the last few years, and they fool idiots into thinking that continuing improvements will be genuinely intelligent. Not possible. The second route might get somewhere eventually, but I would be very surprised to see it in my lifetime. We do not have anything like answers to any of the real questions.
The problem with this is the bank will lose a lot of money if it takes its site offline while it fixes the problem. So when they get mail which could easily be read as 'I have found a security hole in your systems and I could have screwed you royally by messing with it. Give me MONEY!' they, not surprisingly, feel like they're being blackmailed. If the guy had simply sent mail saying, 'you have a security problem, you might want to fix it', then taking him to court would have been stupid. But sometimes someone who thinks they've been helpful and wants money gets unhappy when they're not paid. Squeegee kids at lights have been known to scrape paint off cars that don't pay them occasionally, and the bank no doubt was worried that this guy might go back and exploit the security hole he'd found when he wasn't given money. If you are a major bank, you really can't afford to ignore it, even if only one in a hundred people turned out to be bitter about not being paid, that one in a hundred could seriously mess up the NZ economy for a few days.
Rental is, basically, unenforceable (as in, you always can keep the data if you want to). It is not obsolete.
I frequently rent DVDs via an online company. It's cheap and convenient. I have never ripped one onto my computer, even though I quite easily could. It's not worth the storage space for a movie I most likely won't want to watch again, or at least not at all often.
The difference with music is that you don't generally listen to a good song then not want to listen to it again for at least six months. Music rental never used to be a big business, because everyone knew that if you rented someone an LP or CD it would get copied straight onto tape or computer.
DRM has never been about preventing piracy: if you want to pirate music, you can always connect your sound card's output straight into a recording jack. It's probably easier to do that than to find some clever way to strip the latest DRM off a track anyway. What DRM mainly does is lock consumers into one company's business; if I spend a decent amount of money on iTunes songs, I'm not so likely to go off and buy songs from a competitor because it'll be a pain for me to play them together.
As far as protected music CDs go - this has nothing to do with avoiding piracy. I have a computer, I have good speakers for it. I don't want to waste space or money on a good CD player, and I like to make my own playlists anyway. So I buy CDs with the intent to rip them onto my computer and play them (and not share them - over here this is all perfectly legal). In a few years I'll probably decide I need a new computer, and Dell or someone like that will get my money. Sony will not. This does not make Sony happy. Sony would like me to buy their CD player. So they produce protected CDs, which make it hard for me to listen on my computer.
I'm sorry, but this page is really quite embarassing for the author's parents and any physics teacher's they've ever had. This sort of reminds me of people that read things like A Brief History of Time, a perfectly excellent book, and then try to tell me that the physics is really great and it would be so much better unencumbered by the mathematics.
If you read a paper, you can read each line and verify that it follows from the previous lines, as per the formal definition of a proof (and this is not hugely different from the layman's idea of a massive string of big scary equations, even if your average paper doesn't have any such). This will make it clear to you that the proof works, and you will be happy to say that the theorem is proved (or if there is any error you'll find it) - but you will have no clue _why_ it is true. To understand _why_ a theorem is true you need to be able to read a paper and obtain a mental picture of the situation - which often requires more work than line-by-line verifying the proof, but which often doesn't involve actually checking every statement, and if there is a typo in the proof you may well not catch it. In that sense, physics (or mathematics as your typical undergrad thinks of it) is better without all the `mathematics'.
Not that I'm trying to support the author of the page you refer to, who doesn't seem to either understand or be able to verify the mathematics involved. He certainly hasn't a clue what the Incompleteness Theorem means.
IMO, there is a certain amount of truth behind those pages, in so far as you can construct decent predictive physics models which do not really require space or time. This is fairly obvious, but the author then jumps to `so modern physics must be wrong' which isn't true (you can cast modern physics in such an interpretation without changing any predictions, and if you've any understanding of it this should be obvious as soon as you see the former idea; set an observer to be a point o in a topology, assume a few no-pathological-cases conditions, set information available to the observer to be dependent on the neighbourhoods of o and non-decreasing as you look at smaller neighbourhoods to get the non-spacetime view; this is certainly true of an observer in any standard physics model).
If you think 'rooting a box just to see what it's about' and 'causing $20k worth of damage' are morally equivalent, tell you what - you can have root on this box for $10k, special cheap deal available today.
So criminal punishment has never been about 'an eye for an eye'. Otherwise I'd go robbing banks, safe in the knowledge that whenever I got caught I'd just have to give it back, and sooner or later I'd get away with it. Someone who loses their job might feel aggrieved, sure; that's not an excuse to go breaking stuff. It cost this company $20k, they have every right to recover their $20k. And $20k is easily large enough in most jurisdictions that a jail sentence comes attached to a fine.
Three points. One, when you go to show off a new product at CES it's meant to be new tech, so it probably wasn't built more than a couple of weeks ago at best, at worst it only came out of your best tech wizard's lab last night and can't be turned round or the public will see the components dangling out of the back of the case.
Two, you usually do not get a trial run at these places. Maybe you manage to convince the stuffed shirt who'll be operating it to do a few trial runs back home, but after that you pick up the new box, carry it into a big hall and plug it into a display it's never been near before. Think about what happens when you buy a new DVD player yourself. About one in four times it doesn't work instantly when you plug it into your TV, and you reach round the back, find the loose connection and it works (or something along those lines). You barely even notice the inconvenience at home, being as it took you about a minute to fix and you weren't in a hurry. At CES, that's it, demo failed.
Three, Murphy's Law is strong. There's the Timex story where one in about a dozen durability tests on their watch failed - only the one that failed was the live broadcast one. There's the Win98 issue - OK, that OS did crash a lot, but going by how often it crashed in use, you'd have said that the chance of it BSOD'ing during one short tech demo should have been about one in twenty. Et cetera.
Last week another government official explained that the new licensing laws (extending drinking hours) would probably increase the number of arrests for drunkenness-related crime, and this should be taken as a sign that those laws are a success.
Suppose you had a chain of people 3,000,000km long, and you had them do a Mexican wave. It'd take (a lot) more than 10 seconds to go from one end of the chain to the other because people don't react that fast.
Now suppose you gave each person a Bleepy Thing (tm) which you have sychronised beforehand so they go off at staggered intervals, the last one at the far end of the line 2 seconds after the first. You have the chain of people do its Mexican wave by standing as soon as their Bleepy Thing goes off. Wave velocity will be approximately 5c. There's no problem synchronising the bleepy things, just set them to go off at the right time intervals when they're all together in one place and then move them fairly slowly (like 100km/s is fine) to the right places in the chain.
So why doesn't that break relativity? Answer: the wave does not carry information that fast. In fact the only information you get from the far end of the wave is the time the bleepy things were set to go off at - which reached you much slower than light speed when the bleepy things were sent down the chain beforehand.
This is much the same trick just done with a light wave not a Mexican wave.
Your comment is true if you assume the only way to find collisions is to brute-force search the whole space of all possible messages. It no longer applies when you start using a clever algorithm that narrows the search space; it's quite possible, if very unlikely, that the SHA-1 algorithm cannot find a SHA-1 collision which is also an MD5 collision. For example an algorithm which tries to find collisions in a simple block hash may try to add blocks and preserve the hash; this algorithm obviously never produces a simultaneous collision with the size hash (though of course collisions must exist).
More to the point, while it seems likely that there exists an algorithm capable of finding simultaneous collisions in both MD5 and SHA-1 in about 2^127 time, it's hard to analyse two different hashes together and isn't all that interesting when each hash is separately broken.
Slashdot Mirror
With regards to NVidia - well, what they want in an employee is someone who's not stupid and who is willing to do some work. A lazy genius is no use. That sort of person will get a good GPA. As regards the test thing - well, if you mean they insist you do it and blacklist you if you don't, then that is being a prick. But if they're simply allowing you to do a first screening (and a lot of students will go to a career fair expecting to be there all day and do have time) on the spot rather than waste money going to their HQ half way across the country, I don't see any problem.
For what it's worth, IMO if you interview someone who you'd be happy to work with but lacks technical skills, you probably should tell them - it doesn't take long to give them an idea of where their problems are. If you have someone you think is a prick, then whether or not their technical skills are up to scratch just give them the form letter.
Notes remaining legal tender isn't really all that uncommon. What everyone does is to simply stop reissuing the old variety, so every time an old style note goes into a bank it gets destroyed, and a new style note comes out in its place. Give it six months and you still see a few old style notes about, but not very many - think about how long you normally hang onto a note for: a few days at most. This is also why you don't see too many ripped and tattered notes about, banks destroy them when they get them.
No, of course they can't. The idea is that a lot of developers who don't want to write an Office competitor anyway may well be interested in the freebie: it makes their product look like an MS product and it's (presumably) easier to use the freebie than write your own MS-style UI code. For the average customer, MS-style means it looks professional. Which is a big selling point.
What MS get out of this is that when you have a desktop full of applications which all have one UI style, any other style looks out of place, so the competition looks bad. Same deal as with VB: you can write your app in VB and it does a lot of GUI creation for you, but what you get will be MS-style buttons and so on.
No. OK, so the way fusion is known to work is this: you have nuclei which are moving fast, they smash together and fuse. This gives off energy.
Now, the way this guy's done it (which is not original in method, but technically the guy's got to know what he's doing) is that you use an electric field to accelerate the nuclei. The problem with this is that you can't (easily, anyway) either accelerate large numbers of nuclei or keep the (vast) number which don't fuse from losing most of their energy to the environment. The result is that although the total energy out is larger than the total in, it's only by a very small fraction, and if you try to generate electricity from the heat you end up with less than you used to accelerate the nuclei. Generators aren't 100% efficient (per thermodynamics, they can't be even in theory).
In the Sun, you have a whole lot of high energy nuclei (it's hot near the centre of the Sun), so that when a couple of especially energetic nuclei collide they fuse; when they miss energy isn't dissipated much because they simply bounce off other energetic nuclei. The result is that fusion keeps going without energy needing to be put in all the time. This is where we have a technical problem now: we can't keep fusion going without putting in vast amounts of energy, and until we can reduce the energy input (to at least a sensible percentage of the total energy output, like 50% instead of 99.9%) we won't have any fusion power stations.
Cold fusion is different: the idea is that you can persuade nuclei to fuse when they don't have enough energy for the conventional method. This isn't impossible in theory, but in practice it hasn't been made to work even with lots of energy input.
The other problem with mail votes is the possibility of vote selling. Particularly if you allow this override-later thing.
Suppose you and your wife vote by mail, and then the final count has the people you voted for win by one vote.
You could probably get quite a bit of money by offering to override your votes, on the basis that an unidentifiable masked man made you do it at gunpoint (how can they prove you're lying?).
Override-later still doesn't get rid of the possibility of intimidation, anyway. OK, you can come along later and say someone made you vote this way - but either you let people do that without following up on the accusation, in which case a lot of people will do it because they changed their minds, and potentially some people will do it because they were told to, or alternatively you follow up on accusations, in which case the police probably get swamped with malicious accusations (my vote isn't worth much, and I can use it to get my hated neighbour...) and in the real cases of intimidation no-one changes their vote anyway on pain of broken arms. Same deal with people giving evidence against the Mafia - even with police protection witnesses were kept quite fairly effectively.
The only way you can really avoid intimidation is to have the vote be a one shot, in secret thing, so no-one but you can find out after the fact how you voted. That being the case, since we can now do both an electronic and a paper vote and compare results, it's probably a good idea to do both since it is then much harder to rig the system (and you can announce the electronic count straight away and it isn't much more work than a standard paper vote).
Except that it's very easy to intimidate with vote by mail. You just find someone who is, go round to their house with a gun and make them fill it in as you want it.
Paper ballots have always had problems with ballot box stuffing and so on. Electronic ballots are easily tampered with electronically (and despite what some people on this site seem to think, you cannot build in voter verifiability without losing anonymity).
If you wanted to make it hard to rig an election, IMO the best way to do this is to have an electronic count and a paper count. Use voting machines in polling stations which send the vote to a central computer electronically and print off a paper copy which the voter picks up, verifies and puts in a ballot box themselves. Then you get the results quickly from the electronic system, and the next day you go and count all the paper votes (with scrutineers from all parties) and check that the electronic count and paper count at each polling station match. That way, to fiddle the count you'd have to tamper with the electronic system and with the paper system, and you'd have to get everything to match up. It wouldn't even be much more work than the standard paper votes system.
Not really. You can issue a crypto receipt of vote, fine. Let's assume that the crypto system is unbreakable, since it doesn't really matter. However the voter's arm is not unbreakable, and you have to ensure that someone isn't standing behind the voter either forcing them to vote this way, or after the election forcing them to check their vote and show it's as required.
Even if you can avoid that problem, then you have the problem that you have a central counting machine, which has an encrypted record of the votes. But it cannot add these up unless it can break the crypto - and that would mean that everyone else could, too. So the individual voter can verify that the central machine received his vote correctly, but he cannot verify that it counted the vote correctly. For example, it wouldn't be too hard to write a program which changed the vote counting code so that every 100th Democrat vote was counted as Republican all day, changed it back to the original after voting closed then wiped itself. And you couldn't then discover that without getting everyone to check their votes and adding up the results - and that would mean everyone would know who voted for what.
Godel's theorems have nothing to do with this. You obviously don't understand the meaning of either 'complete' or 'consistent' in the mathematical logic sense - which is NOT the same as the common English sense.
Godel's theorems do not say 'you cannot do this with computers', they say 'you cannot do this'. Whether you have a hunk of dead tree or a hunk of silicon involved is totally irrelevant.
So how do you design an electronic system with individual voters able to verify the total count?
(hint: not possible unless you introduce a 'trusted body' who is allowed to know who everyone voted for)
A paper ballot is obviously proof against electronic tampering. That doesn't make it perfect - after all, if you keep your money stuffed in a mattress it's safer from hackers than if you had it in a bank (as electronic records). But in the former case you will lose everything if a thief comes round.
The problem you have is that you want an election system where only the voter knows how they voted (to avoid intimidation et cetera) but also the counting is verifiably accurate. Unfortunately these aren't compatible. Somewhere there has to be a compromise.
Traditional paper ballots have the problem that you can't detect intelligent tampering with a small number of votes (which can be enough to swing a tight election). Ballot box stuffing, et cetera, are not new.
An intelligently designed electronic system would be reasonably good - i.e. you could easily set up a paper trail which would allow the voter to check that the machine had correctly registered their vote, with not too much trouble you could set up a cryptographic check to ensure that the vote correctly reached a central counting computer. But you couldn't verify that the computer had then performed its job correctly without losing the voter anonymity. Open source counting software would make it slightly harder to fiddle, but it would not be very hard to write a program which runs, fiddles the count then deletes itself afterwards.
The problem with the current US system is that it is not well designed (or at least it's designed by people whose primary concern is profits not a secure system, and they're doing well at that).
There is a difference between 'I want the world to be like this' and 'The world is like this'.
The last time the monarchy in the UK had any power, there was no such thing as the USA. If you want to continue to argue this, you can produce an example of the monarch using their supposed powers any time in the last century or so first, please.
'Inalienable rights' is still just a phrase. Paine's variant (and there have been several) is simply the viewpoint of liberal English society around his time. If what you mean is 'Paine's definition doesn't change over time' then please say so. I'll point out that there are a whole lot of things we'd probably both feel we have a right to which Paine's definitions don't cover. Such as you'd probably object to the police if I set up a giant speaker shouting abuse at you just outside your house. The fact that the police would undoubtedly haul me off for causing a breach of the peace were I to do so is a reflection of the fact that society as a whole does not like people being obnoxious.
Have a look at http://www.spectacle.org/0400/natural.html . It might help you understand the difference between 'right' and 'want' or 'can'.
As far a 'nonhumans cannot think' - well, firstly I see no barrier to recreating a human brain structure by computer simulation (it'd be prohibitively expensive and probably not practically useful but not impossible) and such a thing would certainly pass the Turing test, assuming you didn't pick someone really stupid to copy the brain structure from - secondly, in terms of society, if you prefer you can read 'majority preference of individuals in society' for 'think'. It doesn't affect my argument.
Gordon Brown, who said the thing you originally objected to, is the Chancellor of the Exchequer of the UK. That is not part of the USA. So the US constitution is totally irrelevant.
As regards elections - the details are different, the result is similar. You may have voted as a party member for your political party to put forward a Presidential candidate, but you have never voted directly for a Presidential candidate. You voted for a member of the electoral college who you knew would vote for the Presidential candidate of your choice (and he has the right to change his mind and vote for the other guy, but this almost never happens). In the UK, I vote for an MP who has duties above and beyond getting the guy I want to be prime minister in, but generally people do not vote for an MP because of who he is but for the MP because of the party leader he represents. Put it another way, when did you last have a President who was not the party political choice of one of your two major parties?
Yes, British people are officially subjects of the Queen. Which means nothing in reality. The Queen has no real power, she simply does as told to by the elected government.
As regards 'we made our government protect our rights' - yes, I would really love to be living in a country where the Patriot Acts are well on the way to creating an outright police state, et cetera. Wake up and look at what your goverment is actually doing.
'Inalienable rights of humanity' - this simply translates as 'we can do what society in general thinks is OK'. It sounds good, but your 'inalienable rights' change over time. Pre civil war, every man had the inalienable right to liberty, except if they were black. Post civil war, things were different. If your society decides it doesn't like your inalienable right to religious freedom (many countries around the world) then talking about your 'inalienable rights' will simply make you look pseudo-intellectual as you're dragged to the gallows. Paine wasn't saying 'this is the way things are', he was saying 'this is how things should be'. There is a very big difference.
You know how the UK is not actually part of the USA? That means we do not follow your laws or constitution.
If you want to go on about the UK, please try to understand that.
Furthermore, we have a thing called elections, that being where the guy with the most votes is the prime minister (like the US, really). Except our version of elections doesn't involve whoever is elected receiving huge 'campaign contributions', AKA bribes, from big companies. Which means we are not so badly affected as you are by political lobbyists. Gordon Brown can actually come out and say things like 'the record companies are morally wrong here' and not risk losing any campaign funds for the next election.
If you want to talk about rights - we elect the government, then they pass laws. If we don't like the laws, we elect a government that will repeal those laws. We, the people, decide what rights we think we should have. That is how goverment works: you are not allowed to do something if your fellow people disagree. The alternative is called anarchy, and it does not work in a civilised society (more accurately, if the government were to vanish you would quickly get rule by the strongest, and that eventually would become a democracy in a country with sufficient natural resources).
There's a big difference between 'funded by license fees which the government requires everyone to pay' and 'government controlled'. The BBC is not known for being friendly to the government. If anything, it goes out of its way to demonstrate its independence, it tends to be more attacking than ITV.
VOA _is_ government controlled, it's essentially a news service that's grown up out of a news/propaganda service that was broadcast mainly for the benefit of the Russians.
In any case, if you actually did something like that to a Diebold machine in the US, Diebold would descend upon you with the full force of the DMCA, try to shut you up, and not do a thing about the machines.
Then the Republican Party would happily go on to violate said DMCA by using your vote-rigging code without permission.
"Even complicated proofs involve an extraordinary amount of simple symbol-pushing that a computer could do easily."
Some do, some don't. Yes, there are occasionally (boring) papers publised where someone has simply done a load of algebra to get a result. But most serious mathematics does not work that way. There are a few people trying to formalise proofs so that a computer can at lest verify them, and that hasn't really got very far (a computer can verify simple logic, but a 2 page proof written out in simple logic could fill a book, so no-one does that). Computers can be useful, but mainly for checking lots of simple examples.
You can easily program a computer to perform algebraic manipulations (MAPLE et al.). This is not the same as producing serious proofs. I doubt a computer will be able to replace a professional mathematician any sooner than it will replace a (good) professional translator, or write a good book.
Our brains may well be (probably are) not deterministic, at least in so far as there will be a whole lot of atomic-level effects that most likely do affect our thought processes. That doesn't mean there is no way for a computer to emulate it. There is no reason why you cannot connect a Geiger counter up to a computer and use that to give you any random bits you might want.
The basic problem is that AI researchers go down one of two routes.
One, they try to get something that looks like intelligence, or is actually useful (game AI, say). Which means they do something which is far too rigid for 'genuine AI', whatever that's meant to be. Essentially, a big list of rules plus some kind of selection system involving a bit of (pseudo-)randomness and a bit of scoring by past results, which they call learning.
Two, they try to set up something which might have enough flexibility to include the possibility of genuine thought. They don't know how to teach it (humans do not magically acquire abilities, we learn by example and reward, and how do you simulate that?), they don't know how to set up a structure which can learn (why is one brain better at learning than another? why do we have the ability to learn a complex language when a monkey does not?). In a couple of years their grant runs out and they have nothing resembling a useful result, so no new grant, game over.
The first route will never get anywhere with producing 'intelligence'. It produces stuff like good chess programs, conversation boxes, et cetera; all of these have improved over the last few years, and they fool idiots into thinking that continuing improvements will be genuinely intelligent. Not possible. The second route might get somewhere eventually, but I would be very surprised to see it in my lifetime. We do not have anything like answers to any of the real questions.
The problem with this is the bank will lose a lot of money if it takes its site offline while it fixes the problem. So when they get mail which could easily be read as 'I have found a security hole in your systems and I could have screwed you royally by messing with it. Give me MONEY!' they, not surprisingly, feel like they're being blackmailed. If the guy had simply sent mail saying, 'you have a security problem, you might want to fix it', then taking him to court would have been stupid. But sometimes someone who thinks they've been helpful and wants money gets unhappy when they're not paid. Squeegee kids at lights have been known to scrape paint off cars that don't pay them occasionally, and the bank no doubt was worried that this guy might go back and exploit the security hole he'd found when he wasn't given money. If you are a major bank, you really can't afford to ignore it, even if only one in a hundred people turned out to be bitter about not being paid, that one in a hundred could seriously mess up the NZ economy for a few days.
Rental is, basically, unenforceable (as in, you always can keep the data if you want to). It is not obsolete.
I frequently rent DVDs via an online company. It's cheap and convenient. I have never ripped one onto my computer, even though I quite easily could. It's not worth the storage space for a movie I most likely won't want to watch again, or at least not at all often.
The difference with music is that you don't generally listen to a good song then not want to listen to it again for at least six months. Music rental never used to be a big business, because everyone knew that if you rented someone an LP or CD it would get copied straight onto tape or computer.
DRM has never been about preventing piracy: if you want to pirate music, you can always connect your sound card's output straight into a recording jack. It's probably easier to do that than to find some clever way to strip the latest DRM off a track anyway. What DRM mainly does is lock consumers into one company's business; if I spend a decent amount of money on iTunes songs, I'm not so likely to go off and buy songs from a competitor because it'll be a pain for me to play them together.
As far as protected music CDs go - this has nothing to do with avoiding piracy. I have a computer, I have good speakers for it. I don't want to waste space or money on a good CD player, and I like to make my own playlists anyway. So I buy CDs with the intent to rip them onto my computer and play them (and not share them - over here this is all perfectly legal). In a few years I'll probably decide I need a new computer, and Dell or someone like that will get my money. Sony will not. This does not make Sony happy. Sony would like me to buy their CD player. So they produce protected CDs, which make it hard for me to listen on my computer.
If you read a paper, you can read each line and verify that it follows from the previous lines, as per the formal definition of a proof (and this is not hugely different from the layman's idea of a massive string of big scary equations, even if your average paper doesn't have any such). This will make it clear to you that the proof works, and you will be happy to say that the theorem is proved (or if there is any error you'll find it) - but you will have no clue _why_ it is true. To understand _why_ a theorem is true you need to be able to read a paper and obtain a mental picture of the situation - which often requires more work than line-by-line verifying the proof, but which often doesn't involve actually checking every statement, and if there is a typo in the proof you may well not catch it. In that sense, physics (or mathematics as your typical undergrad thinks of it) is better without all the `mathematics'.
Not that I'm trying to support the author of the page you refer to, who doesn't seem to either understand or be able to verify the mathematics involved. He certainly hasn't a clue what the Incompleteness Theorem means.
IMO, there is a certain amount of truth behind those pages, in so far as you can construct decent predictive physics models which do not really require space or time. This is fairly obvious, but the author then jumps to `so modern physics must be wrong' which isn't true (you can cast modern physics in such an interpretation without changing any predictions, and if you've any understanding of it this should be obvious as soon as you see the former idea; set an observer to be a point o in a topology, assume a few no-pathological-cases conditions, set information available to the observer to be dependent on the neighbourhoods of o and non-decreasing as you look at smaller neighbourhoods to get the non-spacetime view; this is certainly true of an observer in any standard physics model).
If you think 'rooting a box just to see what it's about' and 'causing $20k worth of damage' are morally equivalent, tell you what - you can have root on this box for $10k, special cheap deal available today.
So criminal punishment has never been about 'an eye for an eye'. Otherwise I'd go robbing banks, safe in the knowledge that whenever I got caught I'd just have to give it back, and sooner or later I'd get away with it. Someone who loses their job might feel aggrieved, sure; that's not an excuse to go breaking stuff. It cost this company $20k, they have every right to recover their $20k. And $20k is easily large enough in most jurisdictions that a jail sentence comes attached to a fine.
Three points. One, when you go to show off a new product at CES it's meant to be new tech, so it probably wasn't built more than a couple of weeks ago at best, at worst it only came out of your best tech wizard's lab last night and can't be turned round or the public will see the components dangling out of the back of the case.
Two, you usually do not get a trial run at these places. Maybe you manage to convince the stuffed shirt who'll be operating it to do a few trial runs back home, but after that you pick up the new box, carry it into a big hall and plug it into a display it's never been near before. Think about what happens when you buy a new DVD player yourself. About one in four times it doesn't work instantly when you plug it into your TV, and you reach round the back, find the loose connection and it works (or something along those lines). You barely even notice the inconvenience at home, being as it took you about a minute to fix and you weren't in a hurry. At CES, that's it, demo failed.
Three, Murphy's Law is strong. There's the Timex story where one in about a dozen durability tests on their watch failed - only the one that failed was the live broadcast one. There's the Win98 issue - OK, that OS did crash a lot, but going by how often it crashed in use, you'd have said that the chance of it BSOD'ing during one short tech demo should have been about one in twenty. Et cetera.
Last week another government official explained that the new licensing laws (extending drinking hours) would probably increase the number of arrests for drunkenness-related crime, and this should be taken as a sign that those laws are a success.
Suppose you had a chain of people 3,000,000km long, and you had them do a Mexican wave. It'd take (a lot) more than 10 seconds to go from one end of the chain to the other because people don't react that fast.
Now suppose you gave each person a Bleepy Thing (tm) which you have sychronised beforehand so they go off at staggered intervals, the last one at the far end of the line 2 seconds after the first. You have the chain of people do its Mexican wave by standing as soon as their Bleepy Thing goes off. Wave velocity will be approximately 5c. There's no problem synchronising the bleepy things, just set them to go off at the right time intervals when they're all together in one place and then move them fairly slowly (like 100km/s is fine) to the right places in the chain.
So why doesn't that break relativity? Answer: the wave does not carry information that fast. In fact the only information you get from the far end of the wave is the time the bleepy things were set to go off at - which reached you much slower than light speed when the bleepy things were sent down the chain beforehand.
This is much the same trick just done with a light wave not a Mexican wave.
Your comment is true if you assume the only way to find collisions is to brute-force search the whole space of all possible messages. It no longer applies when you start using a clever algorithm that narrows the search space; it's quite possible, if very unlikely, that the SHA-1 algorithm cannot find a SHA-1 collision which is also an MD5 collision. For example an algorithm which tries to find collisions in a simple block hash may try to add blocks and preserve the hash; this algorithm obviously never produces a simultaneous collision with the size hash (though of course collisions must exist).
More to the point, while it seems likely that there exists an algorithm capable of finding simultaneous collisions in both MD5 and SHA-1 in about 2^127 time, it's hard to analyse two different hashes together and isn't all that interesting when each hash is separately broken.