The problem here is that you're trying to start with DNA forming a complete organism with 'accurate' self-replication. This is hard, yes.
Start simpler. If you look at the citric acid cycle, it is a self-replicating chemical reaction (in the sense that given the building blocks, the main molecule that goes around the cycle ends up doubling itself after one cycle). So all you really need to do to start things off is to randomly generate the citric acid cycle. The various reactions necessary to form the molecules needed for the cycle are a matter of current study, and there are missing points in the chemical theory there, but resolve those points and its a matter of inevitability, not a matter of probability (i.e. you're no longer trying to figure out the chance of this one very rare combination forming, you're just trying to figure out what environment will guarantee the origin of the citric acid cycle within X amount of time)
Now, lets say you don't buy that (there are still those missing reactions/conditions, after all) and you want to start with DNA or RNA. You still don't need a complete organism to take off - you can make do with any sort of analogue of the PCR reaction which amplifies strings currently found in solution. A (very) slow version of this has been observed in water in arctic regions, where the high salinity, low temperature water trapped between a lattice of ice allows the replication to occur without the strands breaking apart over time - this is without the usual PCR enzymes, mind you. You can read that paper at http://www.et1.tu-harburg.de/downloads_et1/ep/publikationen/trinks_ea85.pdf
Additionally, hydrothermal vents could provide an environment perfect for that, since a chamber cycling via thermal convection has been found to be quite good for PCR reactions [Braum, Goddard, and Libchaber "Exponential DNA Replication by Laminar Convection" PRL 91, 2003).
So, while the exact pathway is still a matter of research, I think that this at least shows that the 'critical initial chain' thing may be quite a bit softer of a constraint than you expect.
You can't prove the truth of an axiom, but you can find that some axioms will be inconsistent with experiment. If the universe isn't Lorentz invariant, that will have measureable consequences. It may be difficult to test as perhaps the universe is locally Lorentz invariant but not when you get beyond a certain length or time scale, but it is still in principle something you can test.
The only axioms that will give one trouble is those that do not have measurable consequences. Something like 'The universe we experience is the real one, not just a simulation' doesn't really have a measurable consequence, so its unnecessary for any theory that speaks to experiment. Whether you build the theory with or without that axiom doesn't matter - and if it does matter, you've just found a way to test it.
One can argue about the nature of experiment, of measurement, of reality, etc, but those arguments boil down to asking 'what are we _really_ interested in explaining?'. It seems to me to be getting ahead of yourself to worry about 'what if we're really in vats and this is just hallucination' when you know there's no way of telling and when there are various things that you're able to observe that appear to have patterns. If I have a taste for tea, I put water on the stove, wait till I hear boiling, and pour that over a tea bag because I've seen that process work and I make the assumption that if it has worked once it will work again. Whether or not I'm really tasting it, whether or not the tea really exists or is just a data structure on some computer whose developer is playing the deific version of 'The Sims' doesn't really matter to me;at the end of the processes I can't tell the difference between all of those scenarios but I've still got my tea.
If the tribe only teaches the language to other tribe members, then the only profit Microsoft can make is by selling the version in that language to the tribe. Which means that if the tribe wished to deny them that profit they could simply boycott that version of the product.
If a bank pulled this sort of thing with your finances would you give them a pass because 'it becomes impossible to count _every_ dollar, so our records are differ from the actual amount of money by 3%, but hey, we can probably tell you if that latest check you wrote will bounce or not'? A machine _can_ count every single vote. That's part of the reason to employ electronic voting. If it fails to do that accurately then there is no excuse.
Copying it isn't itself the problem; the problem is copying without attribution. If something is copied without attribution then the work may be credited to the wrong person. It also means that if there were modifications to the work the original cannot be traced for comparison, and if the particular thing copied is only an excerpt it isn't possible to find the context from which that was excerpted without the proper attribution.
The problem with this example is that you didn't use the false part of the assumption to derive the part which works. They're independant. The example you gave can be reduced to the following:
I think X causes sickness. I notice that when I wash myself with water I don't get sick. So I will start washing myself with water.
Notice that the first sentence is totally superfluous. The key bit of reasoning is 'I notice that when I do X, Y happens, so if I want Y to happen I will do X' which is a consistently logical way of approaching the situation.
A better example would be if the belief in demons motivated a local priest to institute regular anointments with holy water. In that case an accidental congruence between the beliefs of the effects of water in one case with the actual effects would promote a successful conclusion. This sort of thing happens when the cause for the analogy being used to solve (or attempt a solution of) the problem is derived in a roughly consistent way with underlying observations, such that it has extensive merit. If I were to take a guess at reconstructing this, it'd be that people observed the importance of clean water for survival long ago, associated clean water with purity, and then that got tied into the religion as a symbol of purification - leading to a useful solution from an incorrect analogy because the analogy respects certain properties of the underlying reality - namely the importance of clean water.
But there are still problems with this. The belief in demons causing illness could promote useless or harmful solutions as well as the accidentally helpful ones. With a flawed base analogy there's really no way to tell which will work and which won't without pairing with observation. Maybe one person would draw the conclusion that exorcisms would help with illness, or that they need to repeat some prayer or phrase until their voice goes hoarse, or whatever.
Part of what science does is it cuts away the parts of an analogy that are inconsistent with the underlying observations until you're aware of the extent that what you're left with is applicable. So you can minimize the number of useless or harmful conclusions and maximize the number of useful ones.
Obviously what the parent meant is that the communication of those bits is speech that must be free.
That has nothing to do with the legality of generating the bits via a certain method, e.g. causing a murder in order to translate images of it to bits.
Whether or not it improves safety is an orthogonal issue to whether or not it creates chilling effects. Something can do both, one, or the other. Having every square inch of the surface of the planet monitored 24/7 might help prevent and punish crimes. But it would also abolish privacy and make everyone feel as if they always had to watch their actions and gauge what they say and do against the reaction of those in power because someone is always watching.
Having a cop every 15 feet standing watch and saying "Move along, citizen." would certainly create an atmosphere of fear and paranoia. People would be afraid to do things which are legal but to which people may object because they're not quite sure IF those actions are legal. People would be afraid to do anything which would draw the cops' ire since when you're being confronted, even if you're legally in the right, you really can't do anything right then and there to defend yourself which doesn't make your situation worse (rather, you have to bring it up in court later when you can't be shot/cuffed/have a charge of resisting arrest added).
Having a CCTV camera with a speaker every 15' gives a similar atmosphere as having a cop every 15' watching 24 hours a day. Maybe the camera can't arrest you on the spot for something, but that doesn't stop the feeling that you might come home to a summons or a fine if you do something that whoever is on the other end of the camera finds objectionable.
Several possible answers. Not sure which one(s) actually apply.
1. An obsession with history. A feeling that its more important what we knew at time X, what we knew at time Y, what we knew at time Z than the actual knowledge.
2. The right way is complicated/mathematically messy and isn't suitable for students first learning to work with the material.
3. The wrong way is a 'close enough' approximation 99% of the time, so if you're training these kids to be engineers who build bridges you don't want the engineering world to waste time and money using relativistic mechanics to figure out the stresses in beams. Gets to 'what is the purpose of learning this'. Is it to spark curiosity and give general understanding about the universe to everyone, or is it to give the tools to build things.
4. Combination of 2 and 3. The wrong way covers most of what's going on and introduces fundamental ideas that don't change when you go to doing things the right way (i.e. conservation laws, Hamiltonian and Lagrangian formalisms, etc), and the wrong way provides a set of easy tools to use to solve many many different problems, which can then be 'patched' in case you want to do things right.
5. The teachers at that level don't know the right way, or don't understand it, so cannot properly teach it.
6. We're just in a process of converting from teaching the old, wrong way into teaching the new, right way (though it seems it's taking > 1 century at this rate).
7. Some places do teach the right way and you were just unlucky and went to a place that subscribed to 1-6.
The way I would parse this is that such a game would ammount to gambling, and would/should/??? be covered by the same laws that cover gambling. Which brings up a question: in a place where gambling is legal, is it legal to bluff in poker?
But aside from that, the thing to look for is: if the company running this thing were to suddenly close up the market, or delete a user's resources, would they be liable for anything? If not, then I don't see there being a similar issue with users playing that game scamming others playing the game if the scam is allowed for by the overseeing company. Because what's basically happening is that the company agrees that a particular set of resource deletions and creations are permissible and that sets the tone for the whole thing.
What will really creep you out is that any sort of induction coil or piezoelectric speaker can also act as a microphone. Of course, it would have to be plugged into an audio input jack on the sound card for that data to actually get anywhere. But just something to think of if the day ever comes when there's an 'sound input/output' jack.
This brings up a point I've never had answered to my satisfaction about the GPL.
Let's say I write a program that incorporates some GPL code and I release it under the public domain. Have I violated the GPL? Or, lets say I release it under the GPL, but then I refuse to pursue any litigation over violation of the license of my code. So someone takes a piece of novel code out of the program and does binary only distribution, etc. Where do the various legal responsibilities fall? Does it end up being that the developer of the very first bit of GPL code that forced everything in the chain to be GPL gets to sue, or can anyone on the chain sue (everyone on the chain?) or no one? Or is it simply that there is no violation of copyright here?
It seems to me that the interpretation would be that that segment of GPL code that had been included remains GPL but the rest can be public domain, in which case its (someone's) responsibility to track the licensing of each line of code separately. But in the second case, it's all GPL (though 95% of it is unenforced GPL) so that license data gets mangled.
This isn't a straw man as it directly responds to the parent's comment "Whats wrong with someone wanting to make a living doing something they enjoy?". It has nothing to do with consideration of rights. I was simply answering that wanting to make a living does not equate to being guaranteed to be able to.
Now, if you want to bring in rights I submit that we'll never find a point of agreement. It's a subjective decision, which set of rights you think is more fundamental. Perhaps you think a person has a fundamental right to control the retransmission of their ideas and creative work. I don't. Is there really anything we can say to eachother beyond that?
Go for it. It doesn't really matter because currently I'm on a grant, so its not really my employer that pays me, its a funding source that said 'yes, we're willing to pay for you to do research because we think its worth being done'. I don't get the funding, it doesn't mean I'll stop working. But it may mean I work on X instead of Y, or I do less of X because I have to work a second job. It seems someone thinks that generating that research is worth the investment, and guess what, its all stuff that basically goes into the public eye. Though I've only got one paper to my name right now as a result of the stuff I've been doing, you and anyone else can go grab it off of arxiv for free.
If at any point in time the people who make my work possible say 'sorry, we don't think your research is worth paying for' then I have to face that. I either have to change the research I do (go for the high-interest grants) or I have to start flipping burgers. It's a reality I'm prepared to deal with. Because I have no right to demand that people fund me to work on some idea of mine that I personally want to work on. I can only say 'I want to work on this, does anyone want to pay me so that I can do it?'. But I have no right to expect to actually receive money for that. And if some competing proposal gets the money and mine doesn't, then I have to live with that.
And I probably still would do the work anyhow, because thats what I want to do. It just means that if I'm getting paid to do it, I can spend my workday on it rather than something else.
The problem with your analysis is that you restrict yourself to point mutations, but this is not the case by far in nature.
In bacteria for example, large segments of their genetic code can be transfered into plasmids which are then copied into other bacteria. So imagine you take your linux kernel example, and there's two versions which have a basic set of drivers A through F, but then version one has driver G and version two has driver H. Because of horizontal gene transfer, you can end up with a single kernel that now has A through H. In effect, a beneficial mutation only has to occur in a single bacterium, and can then be transferred to the rest of the bacterial population without interfering with the innovations that those other bacteria have developed.
Another thing that can happen is that during replication, a bacterium makes a copy of some subsegment of its own genetic code and tacks it on, so it has two copies. Then the second copy can mutate and change function without damaging the function of the first. That is, it decreases the number of harmful mutations because there is redundancy. Imagine it in the linux kernel example - you can make a copy of a function then call both every time that function call appears and as long as one of the two works, the program works (i.e. if one throws an exception or something, it uses the result from the other). You'd suffer a loss of efficiency because of running the code twice but it wouldn't kill the thing and it'd allow a safe zone for incremental improvement.
Yet another thing is, the linux kernel is expressed in a particular language which describes the conversion of information into actions/physical realizations. That language itself is subject to evolutionary forces in biology (at least, when that language first emerged it was), so the language itself can change to bias mutations towards the beneficial. Think of it like evolving in a language like C which has syntax (a mutation will almost always prevent the program from compiling) compared to evolving in a language like assembly where you can make point changes without often breaking the program entirely. Proteins are very flexible as far as switching individual amino acids except at particular active sites, and you can get a continuous change in function. In an abstracted programming language, even something as low level as assembly, there is no continuous change.
So really, it only looks daunting to develop complexity with evolution. The process itself is much deeper than point mutations, and because of this it has far more power to innovate than one might otherwise expect. Especially since evolution operates on its own mechanisms - if a particular transformation of genetic material biases things towards beneficial changes, that transformation can be developed, and you see this. Things like horizontal gene transfer, gene duplication, homologous recombination, sexual crossover, error control in copying, differing mutation rates between organisms, organization of genetic material into chromosomes, and so on.
If this is a serious problem for content creators then they should simply stop creating. Right now, despite piracy, people are STILL making enough money that they'd rather be in that business than flipping burgers and can sustain it. But if they're really hurting, then they should go and flip burgers or become lawyers or whatever so they can make ends meet, rather than expecting the courts to force people to pay for their stuff.
Extracting fines for people seeing information without the creators permission is more theft than copying their information ever was, and I certainly don't want a system where that is their means of income. In my mind, rather than living in a world like that I'd by far prefer all of these injured content creators to simply stop creating content, if it hurts them so much to do so.
Either people will say 'What, no more music? That sucks! Hey, will anyone make a song for me if I give them this donation?' (or groups will do so), or people will say 'Huh... There's less music, but strangely I don't miss it so much, and I can find people who are still willing to make music for free as their hobby or use alternate methods to extract an income from making music.'
Everyone keeps saying 'if we didn't have copyright, there would be no creativity'. Well, I'm saying put it on the line and try it. Speaking for myself, it won't stop me from doing research or writing software if I can't hold a copyright on those things.
Except if you know that opensource/freeware is what you want and you search for it specifically on places like freshmeat and sourceforge. Or just add 'GPL' or 'opensource' to your search.
Not necessarily. One has to consider what information is meaningful in the context of the AI and what information is not meaningful. Take something in which you have a signal that contains certain ordered information and a certain gaussian noise term. You are interested in the ordered information, so extracting that and storing that is lossy - it may reduce the signal from a megabyte to 5 bytes which correspond to five symbols transmitted over the lossy medium. But then even if you write 'add gaussian noise with this standard deviation and that amplitude' you have a lossy compressor because byte-for-byte you get a different signal even if its statistically similar. To make it lossless, you have to include a record of a lot of useless random information which the AI may not care about because it is meaningless in the AI's context.
Every intelligence we know of emerged in an environment where stresses made certain factors important and other factors unimportant. By taking that into account, intelligences can be excellent compressors because they know what to discard. This doesn't even have anything to do with clever decomposition of data into generators - you achieve a lot simply by throwing away what isn't necessary.
Honestly, of the things that bother me about his book, that's not horribly high on the list. The main problem is, nothing he talks about has any ability to predict behavior either qualitatively or quantitatively.
He makes no mention about how to crossover from a microscopic theory to his cellular automata stuff, so even if you can say 'wow, that looks like seashells' when you're presented with some new physical problem you can't just look up his book and figure out what the equivalent CA model would be.
And he doesn't really cover in any depth the intrinsic qualities of doing things with a highly discretized model compared to continuum modelling. Discretization produces a lot of structure in its own right, and that structure may interfere with certain symmetries that the real system has - e.g. translation invariance, invariance when going to a moving frame of reference, and conservation laws. As such, these models tend to give you things which are highly unstable to noise (Conway's game of life collapses to interlocking horizontal and vertical stripe patterns if you add the slightest bit of noise for instance), or work because things exist within the CA model that actually recover those symmetries (but detecting if you're in the first or second case is again something that needs to be covered) - the spiral defect chaos CA models are an example of this, where you essentially recover some sort of continuum dynamics and the discretization doesn't kill it.
I certainly have respect for the method of CA modelling. It can be incredibly useful in simplifying simulation, especially in the case where the full continuum approach ends up having operators which are very numerically unstable. If you can write a simple CA model for resolving the function of those operators, it can save you a ton of time. But I don't really feel that Wolfram's book gives one the necessary set of tools to do that sort of work. Rather, it might just give you an idea to try to do things that way, which is not a bad thing. It just means that even comparing this to peer-reviewed journal publications isn't meaningful because thats not the sort of publication it is. It's more of a halfway-between of a popular science book and a textbook.
This gives me an idea for an interesting analysis. I wonder if you take all the weather predictions for the last 20 years or so and compare with the actual weather, if you'd see any patterns when you plot a map of the error as a function of location (and perhaps isolate it to the weather during a particular time of year). If there are particular locations which end up being tipping points, then that tells you something about the dynamics and where you need the highest resolution when you're building your models.
Actually, all we need to know is the ratio of counterfeit to genuine systems and assume that all counterfeit systems are flagged as such. We know that x% of genuine systems is flagged as counterfeit, and that x*(# genuine) = 0.25*(# counterfeit), so the percentage of genuine systems flagged is one quarter of the ratio of counterfeit to genuine systems. Additionally, this lets us put a bound on the number of detectably counterfeit systems out there. There cannot be more than four times as many (detectable) counterfeit systems out there than genuine systems - if that were the case, then 100% of genuine systems would be flagged as counterfeit.
Ah, but there is absolutely nothing stopping companies from dedicating resources towards engineering new needs. Advertising is a huge example of this. It's the same sort of thing the government does, except the wording is a bit different.
Company: You need accountability and support from us. Pay us $500 so we're a scapegoat if something goes wrong, rather than $20 for that repair place who we haven't given our blessing to. Remember, if you go to the $20 repair place, you could get fired, but no one gets fired for going with the official solution!
Government: You need us to protect you from drugs, terrorists, and offensive video games. Pay us to waste a bunch of money on half-baked initiatives to do these things and to cover our new expanded salaries. Remember, if you don't pay us, you're responsible for when anything bad happens to anyone else, cause y'know, maybe if you had paid us we could have possibly with some chance maybe done something to prevent it (but if we don't prevent it, it was because someone else didn't pay us or because we're not getting paid enough!)
It's not a problem of the state. It's a problem with authority. In the absence of a government, all it takes to form one is a group of people getting together and deciding to act together. If that group happens to not have your best interests in mind, you're back to where you started. Even if they do, maybe their interpretation of 'your best interests' isn't the same as yours.
So I ask, and this is a yet-unanswered question to me, what is the absolute minimum government necessary to prevent this sort of nucleation-of-groupings? That is, how do you prevent the formation of large concentrations of power over others without also hindering people's ability to live freely and pursue those things they find worthwhile (including the ability to assemble and group together to accomplish things).
Slashdot Mirror
The problem here is that you're trying to start with DNA forming a complete organism with 'accurate' self-replication. This is hard, yes.
Start simpler. If you look at the citric acid cycle, it is a self-replicating chemical reaction (in the sense that given the building blocks, the main molecule that goes around the cycle ends up doubling itself after one cycle). So all you really need to do to start things off is to randomly generate the citric acid cycle. The various reactions necessary to form the molecules needed for the cycle are a matter of current study, and there are missing points in the chemical theory there, but resolve those points and its a matter of inevitability, not a matter of probability (i.e. you're no longer trying to figure out the chance of this one very rare combination forming, you're just trying to figure out what environment will guarantee the origin of the citric acid cycle within X amount of time)
Now, lets say you don't buy that (there are still those missing reactions/conditions, after all) and you want to start with DNA or RNA. You still don't need a complete organism to take off - you can make do with any sort of analogue of the PCR reaction which amplifies strings currently found in solution. A (very) slow version of this has been observed in water in arctic regions, where the high salinity, low temperature water trapped between a lattice of ice allows the replication to occur without the strands breaking apart over time - this is without the usual PCR enzymes, mind you. You can read that paper at http://www.et1.tu-harburg.de/downloads_et1/ep/publikationen/trinks_ea85.pdf
Additionally, hydrothermal vents could provide an environment perfect for that, since a chamber cycling via thermal convection has been found to be quite good for PCR reactions [Braum, Goddard, and Libchaber "Exponential DNA Replication by Laminar Convection" PRL 91, 2003).
So, while the exact pathway is still a matter of research, I think that this at least shows that the 'critical initial chain' thing may be quite a bit softer of a constraint than you expect.
Is one allowed to submit tap water/sea water/etc for use in this test and/or have you perform the test at a different ambient temperature than 1 atm?
You can't prove the truth of an axiom, but you can find that some axioms will be inconsistent with experiment. If the universe isn't Lorentz invariant, that will have measureable consequences. It may be difficult to test as perhaps the universe is locally Lorentz invariant but not when you get beyond a certain length or time scale, but it is still in principle something you can test.
The only axioms that will give one trouble is those that do not have measurable consequences. Something like 'The universe we experience is the real one, not just a simulation' doesn't really have a measurable consequence, so its unnecessary for any theory that speaks to experiment. Whether you build the theory with or without that axiom doesn't matter - and if it does matter, you've just found a way to test it.
One can argue about the nature of experiment, of measurement, of reality, etc, but those arguments boil down to asking 'what are we _really_ interested in explaining?'. It seems to me to be getting ahead of yourself to worry about 'what if we're really in vats and this is just hallucination' when you know there's no way of telling and when there are various things that you're able to observe that appear to have patterns. If I have a taste for tea, I put water on the stove, wait till I hear boiling, and pour that over a tea bag because I've seen that process work and I make the assumption that if it has worked once it will work again. Whether or not I'm really tasting it, whether or not the tea really exists or is just a data structure on some computer whose developer is playing the deific version of 'The Sims' doesn't really matter to me;at the end of the processes I can't tell the difference between all of those scenarios but I've still got my tea.
If the tribe only teaches the language to other tribe members, then the only profit Microsoft can make is by selling the version in that language to the tribe. Which means that if the tribe wished to deny them that profit they could simply boycott that version of the product.
If a bank pulled this sort of thing with your finances would you give them a pass because 'it becomes impossible to count _every_ dollar, so our records are differ from the actual amount of money by 3%, but hey, we can probably tell you if that latest check you wrote will bounce or not'? A machine _can_ count every single vote. That's part of the reason to employ electronic voting. If it fails to do that accurately then there is no excuse.
Copying it isn't itself the problem; the problem is copying without attribution. If something is copied without attribution then the work may be credited to the wrong person. It also means that if there were modifications to the work the original cannot be traced for comparison, and if the particular thing copied is only an excerpt it isn't possible to find the context from which that was excerpted without the proper attribution.
The problem with this example is that you didn't use the false part of the assumption to derive the part which works. They're independant. The example you gave can be reduced to the following:
I think X causes sickness. I notice that when I wash myself with water I don't get sick. So I will start washing myself with water.
Notice that the first sentence is totally superfluous. The key bit of reasoning is 'I notice that when I do X, Y happens, so if I want Y to happen I will do X' which is a consistently logical way of approaching the situation.
A better example would be if the belief in demons motivated a local priest to institute regular anointments with holy water. In that case an accidental congruence between the beliefs of the effects of water in one case with the actual effects would promote a successful conclusion. This sort of thing happens when the cause for the analogy being used to solve (or attempt a solution of) the problem is derived in a roughly consistent way with underlying observations, such that it has extensive merit. If I were to take a guess at reconstructing this, it'd be that people observed the importance of clean water for survival long ago, associated clean water with purity, and then that got tied into the religion as a symbol of purification - leading to a useful solution from an incorrect analogy because the analogy respects certain properties of the underlying reality - namely the importance of clean water.
But there are still problems with this. The belief in demons causing illness could promote useless or harmful solutions as well as the accidentally helpful ones. With a flawed base analogy there's really no way to tell which will work and which won't without pairing with observation. Maybe one person would draw the conclusion that exorcisms would help with illness, or that they need to repeat some prayer or phrase until their voice goes hoarse, or whatever.
Part of what science does is it cuts away the parts of an analogy that are inconsistent with the underlying observations until you're aware of the extent that what you're left with is applicable. So you can minimize the number of useless or harmful conclusions and maximize the number of useful ones.
Obviously what the parent meant is that the communication of those bits is speech that must be free.
That has nothing to do with the legality of generating the bits via a certain method, e.g. causing a murder in order to translate images of it to bits.
Whether or not it improves safety is an orthogonal issue to whether or not it creates chilling effects. Something can do both, one, or the other. Having every square inch of the surface of the planet monitored 24/7 might help prevent and punish crimes. But it would also abolish privacy and make everyone feel as if they always had to watch their actions and gauge what they say and do against the reaction of those in power because someone is always watching.
Having a cop every 15 feet standing watch and saying "Move along, citizen." would certainly create an atmosphere of fear and paranoia. People would be afraid to do things which are legal but to which people may object because they're not quite sure IF those actions are legal. People would be afraid to do anything which would draw the cops' ire since when you're being confronted, even if you're legally in the right, you really can't do anything right then and there to defend yourself which doesn't make your situation worse (rather, you have to bring it up in court later when you can't be shot/cuffed/have a charge of resisting arrest added).
Having a CCTV camera with a speaker every 15' gives a similar atmosphere as having a cop every 15' watching 24 hours a day. Maybe the camera can't arrest you on the spot for something, but that doesn't stop the feeling that you might come home to a summons or a fine if you do something that whoever is on the other end of the camera finds objectionable.
Several possible answers. Not sure which one(s) actually apply.
1. An obsession with history. A feeling that its more important what we knew at time X, what we knew at time Y, what we knew at time Z than the actual knowledge.
2. The right way is complicated/mathematically messy and isn't suitable for students first learning to work with the material.
3. The wrong way is a 'close enough' approximation 99% of the time, so if you're training these kids to be engineers who build bridges you don't want the engineering world to waste time and money using relativistic mechanics to figure out the stresses in beams. Gets to 'what is the purpose of learning this'. Is it to spark curiosity and give general understanding about the universe to everyone, or is it to give the tools to build things.
4. Combination of 2 and 3. The wrong way covers most of what's going on and introduces fundamental ideas that don't change when you go to doing things the right way (i.e. conservation laws, Hamiltonian and Lagrangian formalisms, etc), and the wrong way provides a set of easy tools to use to solve many many different problems, which can then be 'patched' in case you want to do things right.
5. The teachers at that level don't know the right way, or don't understand it, so cannot properly teach it.
6. We're just in a process of converting from teaching the old, wrong way into teaching the new, right way (though it seems it's taking > 1 century at this rate).
7. Some places do teach the right way and you were just unlucky and went to a place that subscribed to 1-6.
The way I would parse this is that such a game would ammount to gambling, and would/should/??? be covered by the same laws that cover gambling. Which brings up a question: in a place where gambling is legal, is it legal to bluff in poker?
But aside from that, the thing to look for is: if the company running this thing were to suddenly close up the market, or delete a user's resources, would they be liable for anything? If not, then I don't see there being a similar issue with users playing that game scamming others playing the game if the scam is allowed for by the overseeing company. Because what's basically happening is that the company agrees that a particular set of resource deletions and creations are permissible and that sets the tone for the whole thing.
What will really creep you out is that any sort of induction coil or piezoelectric speaker can also act as a microphone. Of course, it would have to be plugged into an audio input jack on the sound card for that data to actually get anywhere. But just something to think of if the day ever comes when there's an 'sound input/output' jack.
This brings up a point I've never had answered to my satisfaction about the GPL.
Let's say I write a program that incorporates some GPL code and I release it under the public domain. Have I violated the GPL? Or, lets say I release it under the GPL, but then I refuse to pursue any litigation over violation of the license of my code. So someone takes a piece of novel code out of the program and does binary only distribution, etc. Where do the various legal responsibilities fall? Does it end up being that the developer of the very first bit of GPL code that forced everything in the chain to be GPL gets to sue, or can anyone on the chain sue (everyone on the chain?) or no one? Or is it simply that there is no violation of copyright here?
It seems to me that the interpretation would be that that segment of GPL code that had been included remains GPL but the rest can be public domain, in which case its (someone's) responsibility to track the licensing of each line of code separately. But in the second case, it's all GPL (though 95% of it is unenforced GPL) so that license data gets mangled.
This isn't a straw man as it directly responds to the parent's comment "Whats wrong with someone wanting to make a living doing something they enjoy?". It has nothing to do with consideration of rights. I was simply answering that wanting to make a living does not equate to being guaranteed to be able to.
Now, if you want to bring in rights I submit that we'll never find a point of agreement. It's a subjective decision, which set of rights you think is more fundamental. Perhaps you think a person has a fundamental right to control the retransmission of their ideas and creative work. I don't. Is there really anything we can say to eachother beyond that?
Go for it. It doesn't really matter because currently I'm on a grant, so its not really my employer that pays me, its a funding source that said 'yes, we're willing to pay for you to do research because we think its worth being done'. I don't get the funding, it doesn't mean I'll stop working. But it may mean I work on X instead of Y, or I do less of X because I have to work a second job. It seems someone thinks that generating that research is worth the investment, and guess what, its all stuff that basically goes into the public eye. Though I've only got one paper to my name right now as a result of the stuff I've been doing, you and anyone else can go grab it off of arxiv for free.
If at any point in time the people who make my work possible say 'sorry, we don't think your research is worth paying for' then I have to face that. I either have to change the research I do (go for the high-interest grants) or I have to start flipping burgers. It's a reality I'm prepared to deal with. Because I have no right to demand that people fund me to work on some idea of mine that I personally want to work on. I can only say 'I want to work on this, does anyone want to pay me so that I can do it?'. But I have no right to expect to actually receive money for that. And if some competing proposal gets the money and mine doesn't, then I have to live with that.
And I probably still would do the work anyhow, because thats what I want to do. It just means that if I'm getting paid to do it, I can spend my workday on it rather than something else.
They aren't entitled to success. If people don't care enough enable that success, they have no right to still receive that money.
Doing what you most want to do for a living may simply be untenable. Fact of life.
The problem with your analysis is that you restrict yourself to point mutations, but this is not the case by far in nature.
In bacteria for example, large segments of their genetic code can be transfered into plasmids which are then copied into other bacteria. So imagine you take your linux kernel example, and there's two versions which have a basic set of drivers A through F, but then version one has driver G and version two has driver H. Because of horizontal gene transfer, you can end up with a single kernel that now has A through H. In effect, a beneficial mutation only has to occur in a single bacterium, and can then be transferred to the rest of the bacterial population without interfering with the innovations that those other bacteria have developed.
Another thing that can happen is that during replication, a bacterium makes a copy of some subsegment of its own genetic code and tacks it on, so it has two copies. Then the second copy can mutate and change function without damaging the function of the first. That is, it decreases the number of harmful mutations because there is redundancy. Imagine it in the linux kernel example - you can make a copy of a function then call both every time that function call appears and as long as one of the two works, the program works (i.e. if one throws an exception or something, it uses the result from the other). You'd suffer a loss of efficiency because of running the code twice but it wouldn't kill the thing and it'd allow a safe zone for incremental improvement.
Yet another thing is, the linux kernel is expressed in a particular language which describes the conversion of information into actions/physical realizations. That language itself is subject to evolutionary forces in biology (at least, when that language first emerged it was), so the language itself can change to bias mutations towards the beneficial. Think of it like evolving in a language like C which has syntax (a mutation will almost always prevent the program from compiling) compared to evolving in a language like assembly where you can make point changes without often breaking the program entirely. Proteins are very flexible as far as switching individual amino acids except at particular active sites, and you can get a continuous change in function. In an abstracted programming language, even something as low level as assembly, there is no continuous change.
So really, it only looks daunting to develop complexity with evolution. The process itself is much deeper than point mutations, and because of this it has far more power to innovate than one might otherwise expect. Especially since evolution operates on its own mechanisms - if a particular transformation of genetic material biases things towards beneficial changes, that transformation can be developed, and you see this. Things like horizontal gene transfer, gene duplication, homologous recombination, sexual crossover, error control in copying, differing mutation rates between organisms, organization of genetic material into chromosomes, and so on.
If this is a serious problem for content creators then they should simply stop creating. Right now, despite piracy, people are STILL making enough money that they'd rather be in that business than flipping burgers and can sustain it. But if they're really hurting, then they should go and flip burgers or become lawyers or whatever so they can make ends meet, rather than expecting the courts to force people to pay for their stuff.
Extracting fines for people seeing information without the creators permission is more theft than copying their information ever was, and I certainly don't want a system where that is their means of income. In my mind, rather than living in a world like that I'd by far prefer all of these injured content creators to simply stop creating content, if it hurts them so much to do so.
Either people will say 'What, no more music? That sucks! Hey, will anyone make a song for me if I give them this donation?' (or groups will do so), or people will say 'Huh... There's less music, but strangely I don't miss it so much, and I can find people who are still willing to make music for free as their hobby or use alternate methods to extract an income from making music.'
Everyone keeps saying 'if we didn't have copyright, there would be no creativity'. Well, I'm saying put it on the line and try it. Speaking for myself, it won't stop me from doing research or writing software if I can't hold a copyright on those things.
Except if you know that opensource/freeware is what you want and you search for it specifically on places like freshmeat and sourceforge. Or just add 'GPL' or 'opensource' to your search.
Not necessarily. One has to consider what information is meaningful in the context of the AI and what information is not meaningful. Take something in which you have a signal that contains certain ordered information and a certain gaussian noise term. You are interested in the ordered information, so extracting that and storing that is lossy - it may reduce the signal from a megabyte to 5 bytes which correspond to five symbols transmitted over the lossy medium. But then even if you write 'add gaussian noise with this standard deviation and that amplitude' you have a lossy compressor because byte-for-byte you get a different signal even if its statistically similar. To make it lossless, you have to include a record of a lot of useless random information which the AI may not care about because it is meaningless in the AI's context.
Every intelligence we know of emerged in an environment where stresses made certain factors important and other factors unimportant. By taking that into account, intelligences can be excellent compressors because they know what to discard. This doesn't even have anything to do with clever decomposition of data into generators - you achieve a lot simply by throwing away what isn't necessary.
Honestly, of the things that bother me about his book, that's not horribly high on the list. The main problem is, nothing he talks about has any ability to predict behavior either qualitatively or quantitatively.
He makes no mention about how to crossover from a microscopic theory to his cellular automata stuff, so even if you can say 'wow, that looks like seashells' when you're presented with some new physical problem you can't just look up his book and figure out what the equivalent CA model would be.
And he doesn't really cover in any depth the intrinsic qualities of doing things with a highly discretized model compared to continuum modelling. Discretization produces a lot of structure in its own right, and that structure may interfere with certain symmetries that the real system has - e.g. translation invariance, invariance when going to a moving frame of reference, and conservation laws. As such, these models tend to give you things which are highly unstable to noise (Conway's game of life collapses to interlocking horizontal and vertical stripe patterns if you add the slightest bit of noise for instance), or work because things exist within the CA model that actually recover those symmetries (but detecting if you're in the first or second case is again something that needs to be covered) - the spiral defect chaos CA models are an example of this, where you essentially recover some sort of continuum dynamics and the discretization doesn't kill it.
I certainly have respect for the method of CA modelling. It can be incredibly useful in simplifying simulation, especially in the case where the full continuum approach ends up having operators which are very numerically unstable. If you can write a simple CA model for resolving the function of those operators, it can save you a ton of time. But I don't really feel that Wolfram's book gives one the necessary set of tools to do that sort of work. Rather, it might just give you an idea to try to do things that way, which is not a bad thing. It just means that even comparing this to peer-reviewed journal publications isn't meaningful because thats not the sort of publication it is. It's more of a halfway-between of a popular science book and a textbook.
This gives me an idea for an interesting analysis. I wonder if you take all the weather predictions for the last 20 years or so and compare with the actual weather, if you'd see any patterns when you plot a map of the error as a function of location (and perhaps isolate it to the weather during a particular time of year). If there are particular locations which end up being tipping points, then that tells you something about the dynamics and where you need the highest resolution when you're building your models.
Probably someone has already done this though.
Actually, all we need to know is the ratio of counterfeit to genuine systems and assume that all counterfeit systems are flagged as such. We know that x% of genuine systems is flagged as counterfeit, and that x*(# genuine) = 0.25*(# counterfeit), so the percentage of genuine systems flagged is one quarter of the ratio of counterfeit to genuine systems. Additionally, this lets us put a bound on the number of detectably counterfeit systems out there. There cannot be more than four times as many (detectable) counterfeit systems out there than genuine systems - if that were the case, then 100% of genuine systems would be flagged as counterfeit.
Ah, but there is absolutely nothing stopping companies from dedicating resources towards engineering new needs. Advertising is a huge example of this. It's the same sort of thing the government does, except the wording is a bit different.
Company: You need accountability and support from us. Pay us $500 so we're a scapegoat if something goes wrong, rather than $20 for that repair place who we haven't given our blessing to. Remember, if you go to the $20 repair place, you could get fired, but no one gets fired for going with the official solution!
Government: You need us to protect you from drugs, terrorists, and offensive video games. Pay us to waste a bunch of money on half-baked initiatives to do these things and to cover our new expanded salaries. Remember, if you don't pay us, you're responsible for when anything bad happens to anyone else, cause y'know, maybe if you had paid us we could have possibly with some chance maybe done something to prevent it (but if we don't prevent it, it was because someone else didn't pay us or because we're not getting paid enough!)
It's not a problem of the state. It's a problem with authority. In the absence of a government, all it takes to form one is a group of people getting together and deciding to act together. If that group happens to not have your best interests in mind, you're back to where you started. Even if they do, maybe their interpretation of 'your best interests' isn't the same as yours.
So I ask, and this is a yet-unanswered question to me, what is the absolute minimum government necessary to prevent this sort of nucleation-of-groupings? That is, how do you prevent the formation of large concentrations of power over others without also hindering people's ability to live freely and pursue those things they find worthwhile (including the ability to assemble and group together to accomplish things).