It doesn't matter whether people are "responsible" for their actions. Punishment or reward make sense only insofar as they result in improving our lives. The point of punishment should not be "you knowingly did wrong, therefore I will punish you". The point is "if people who do X get punished, then people are less likely to do X".
Calling it "holding people accountable" encourages sloppy thinking. (For example, it may lead to punishing people for activities that punishment doesn't dissuade, only causing unnecessary pain. Or it may cause people to foolishly prefer punishment for misbehavior over eliminating societal factors that lead to the misbehavior.)
Calling it a disincentive for undesired behavior sets you up to make good decisions about when, where, how much, and when to use an alternative.
Mathematics is not a science - it's its own thing.
In math, you can prove a theory (theorem) is true: a proof in mathematics is a set of typographical transformations using the rules of a system of mathematics that take you from the lemma in your mathematical system to the thing you're trying to prove. It can be verified by a machine with a VERY high degree of confidence.
In science, "proving" a theory means you've given a repeatable experiment whose results strongly support your theory. You can never prove something is right, you can only disprove by finding a counterexample.
I agree that it's just speculation as to the practicality of this thing, but I disagree strongly re: consciousness.
The body is controlled by nerve impulses, which in the end can be measured in simple volumes of chemicals passing through membranes. If the simulation produces the same simulated volumes as the real thing (within some acceptable tolerance of variation), then it will behave like the real thing.
Which means if you explain a concept to it, then ask it to use the concept in a sentence or in describing a design, it will do the same thing (more or less) as the original system - use the new concept intelligently. If you ask it if it's conscious, it will say yes (assuming the original would).
There is no more reason to doubt whether it's "really" conscious than there is for me to doubt you're conscious.
I have a BS in Physics (with some emphasis on quantum mechanics in my coursework and independent reading), and I know of no QM process that we can't simulate classically. It may be (and often is) much slower, in fact it is often impossible for any classical system to be as fast as the QM system, but the simulation can be as accurate as our understanding of the process and our computing power allows.
If we can simulate the behavior of each individual neuron to acceptable tolerance, then simulating a brain is just a matter of simulating them all at once. The only caveat I know of to that is if there was somehow quantum interaction among multiple neurons that caused them to interact as if the (classical) whole were more than the sum of the (classical) parts, but for that to happen there would have to be quantum coherence on a scale of multiple neurons, and quantum coherence on that scale is very delicate and rare - the only examples I know of happen in a vacuum, barring superfluids, superconductivity, and bose-einstein condensate (which occur almost exclusively at temperatures near absolute zero).
I agree with you that something like this is the most likely way to make simulating a brain work, but...
There are very interesting ethical issues here. If you simulate my brain but with some errors that make it batshit nuts, or leave it writhing in agony, are you torturing someone? Does turning the program off make everything OK?
How many simulated people do you torture, then terminate while trying to get it right? Why should it matter if the consciousness of the critter tortured is based on physical interactions of atoms or on the flow of electrons?
I'm agnostic, and I *am* picking a side and standing up for it.
'Agnostic' doesn't mean "I don't know whether to believe in the Christian god" (as your comment re: Zeus appears to imply). Agnosticism can mean either "I don't know whether there's a god or not" or "I believe it's unknowable whether there's a god or not".
A god could take many, many forms. I believe there's strong evidence against a god who created humanity in its image and for the purpose of exalting it.
Not having heard any reasonable theory of the origin of the stuff in the universe, the space-time in which it sits, and the physical laws governing that stuff, I don't know where it comes from. I find it equally hard to believe that some entity outside of the bounds of physical laws created it or that it has no origin. (Giving a physical explanation of its origin just begs the question of the origin of the physics of the explanation).
I don't think you know either. So I call on you, stop pretending you're omniscient: admit you're not atheist, but rather agnostic.
Either your passwords are stored on your harddisk or you're going to have to type them in at a later point. In both cases software is going to be able to get your passwords. And they have that they can get in without a problem, regardless of protocol used.
Unless and until someone develops all these systems (starting with file transfer) to operate on challenge & response using pub/priv key systems. Hmm, in fact I suspect ssh does work that way when you use keys. I need to look into that...
Anyway, if you use such a system, then you can have a token (physical key) which contains your private key in an inaccessible way, but can provide the public key on demand or can be asked to sign a piece of text using your private key. Then the authenticating system sends you random text, your token signs it & sends it back, and the authenticating system validates the signature using your public key.
Then, regardless of the security of the system you're using, your identity is secure. Someone may steal your session and do Bad Things, but they can't ever login again as you once you're logged out on that session. For very dangerous operations, the authenticating system just requests you to authorize the operation explicitly, so even session hijacking won't let an attacker screw you. If necessary your token can provide a way to display a simple message about what operation you're authenticating by pressing OK.
Ideally, the physical key has a way for you to enter a password directly into it. Then someone would have to not only steal your key, but also learn your password to be able to hijack your identity.
If you need to invalidate an old token and migrate to a new one, you can allow a request signed by some authority to invalidate the old key and inject a new one. If you don't like central authorities, web of trust can be used in the same way.
Weight gain is *mass* in minus *mass* out. Now, how much of the mass of what you eat sticks with you may be related to the calories in it. In fact, it seems downright likely. But it is not the end-all.
This notion of "duh, OBVIOUSLY if you consume more calories than you burn, you'll gain weight" is pseudoscientific thinking. Calories are a measure of the chemical energy in your food. Whether your body can access that energy, how it stores it, how the food affects your metabolism, and a whole host of other things will affect weight gain.
I have no special training in food science or metabolism, but the "obvious" platitudes spouted off here wrt weight gain are obviously wrong.
I agree that he should carry through the "self-reported" modifier when he talks about the group with his preferred response, but the excerpt you gave doesn't have any problem in the science. He isn't saying the results are from people who are actually good at math - it clearly says "self-reported".
Cut-n-paste: Factoring in all these issues, various groups have attempted to calculate a true economic cost for electricity generated by the most modern designs proposed. Because if an actual cost per kW*h can be calculated, then it is possible to compare it to other power sources to determine if such an investment is economically sound.
In 2003, the Massachusetts Institute of Technology (MIT) issued a report entitled, "The Future of Nuclear Power". They estimated that new nuclear power in the US would cost 6.7 cents per kW*h.[1] However, the Energy Policy Act of 2005 includes a tax credit that should reduce that cost slightly.
The lifetime cost of new generating capacity in the United States was estimated in 2006 by the U.S. government (the 2007 report did not estimate costs). Nuclear power was estimated at 5.93 cents per kW*h. However, the "total overnight cost" for new nuclear was assumed to be $1,984 per kWe - as seen above in Capital Costs, this figure is subject to debate.
A 2008 study based on historical outcomes in the U.S. said costs for nuclear power can be expected to run $0.25-.30 per kW*h.
A 2008 study concluded that if carbon capture and storage was required then nuclear power would be the cheapest source of electricity even at $4,038/kW in overnight capital cost.
In 2009, MIT updated its 2003 study, concluding that inflation and rising construction costs had increased the overnight cost of nuclear power plants to about $4,000/kWe, and thus increased the power cost to 8.4c/kW*h.
Interesting - the fellow I know is an extreme christian who actually avoids learning about quantum mechanics, because he doesn't want it to be true (?!), I'm not sure of his take on big bang, he believes in limited evolution (some speciation, but not that mammals and reptiles have a common ancestor, nor the animals having a common ancestor with man).
I know him from working on software dev projects with him, but he is one of my few work friends who are also friends outside of work. He and I are both National Merit Scholars; his background is in marine biology.
He was not a christian until probably his 30s - his wife "turned" him. Now he is a deacon of his church.
I actually have a Neo 1973 - no need for the capacitor, nor is there a need for it in the newest Freerunners. I do agree that he should read up on it and hang out in #openmoko on irc.freenode.net to see what to expect. Early on, software was very buggy. Now, afaik, the only persistent problem is short battery life (about a day with normal usage).
"it hardly works" is inaccurate. There are issues he should understand *is* accurate.
Get a Freerunner. It's an open phone with: GPS, internet via gprs, accelerometers, full bluetooth, wifi, 640x480 touchscreen. It runs any of several flavors of Linux (including Debian or Android, but my personal choice is SHR) and there are already Free gps programs that use OpenStreetMap (TangoGPS or Navit).
It's about $250, IIRC, but of course you can carry it around and use it to browse the web and receive calls, as well as using it in-vehicle for navigation.
As the AC said, the chemistry doesn't matter. Energy is mass; it's just hard to convert significant amounts of one to the other. It doesn't matter if the energy is chemical, nuclear, kinetic, gravitational.
It's just (much) easier to see the change in mass in a nuclear reaction than in a chemical one.
BTW, if the released energy from the chemical reaction all went to heat in the solution, none of which escapes, then the mass of the solution wouldn't change. All that matters (no pun intended) is how much energy enters or leaves the system - there is a corresponding change in mass.
Slashdot Mirror
Creating a new simulated neuron in a simulated brain is trivial...
So what?
It doesn't matter whether people are "responsible" for their actions. Punishment or reward make sense only insofar as they result in improving our lives. The point of punishment should not be "you knowingly did wrong, therefore I will punish you". The point is "if people who do X get punished, then people are less likely to do X".
Calling it "holding people accountable" encourages sloppy thinking. (For example, it may lead to punishing people for activities that punishment doesn't dissuade, only causing unnecessary pain. Or it may cause people to foolishly prefer punishment for misbehavior over eliminating societal factors that lead to the misbehavior.)
Calling it a disincentive for undesired behavior sets you up to make good decisions about when, where, how much, and when to use an alternative.
Or a frequently wrong man.
Mathematics is not a science - it's its own thing.
In math, you can prove a theory (theorem) is true: a proof in mathematics is a set of typographical transformations using the rules of a system of mathematics that take you from the lemma in your mathematical system to the thing you're trying to prove. It can be verified by a machine with a VERY high degree of confidence.
In science, "proving" a theory means you've given a repeatable experiment whose results strongly support your theory. You can never prove something is right, you can only disprove by finding a counterexample.
In British English, "the group are", not "the group is". In American English, it's as you say.
Really? I'd like to see a citation.
And even I at 14 could see that Omni magazine was mostly full of shit.
I agree that it's just speculation as to the practicality of this thing, but I disagree strongly re: consciousness.
The body is controlled by nerve impulses, which in the end can be measured in simple volumes of chemicals passing through membranes. If the simulation produces the same simulated volumes as the real thing (within some acceptable tolerance of variation), then it will behave like the real thing.
Which means if you explain a concept to it, then ask it to use the concept in a sentence or in describing a design, it will do the same thing (more or less) as the original system - use the new concept intelligently. If you ask it if it's conscious, it will say yes (assuming the original would).
There is no more reason to doubt whether it's "really" conscious than there is for me to doubt you're conscious.
I have a BS in Physics (with some emphasis on quantum mechanics in my coursework and independent reading), and I know of no QM process that we can't simulate classically. It may be (and often is) much slower, in fact it is often impossible for any classical system to be as fast as the QM system, but the simulation can be as accurate as our understanding of the process and our computing power allows.
If we can simulate the behavior of each individual neuron to acceptable tolerance, then simulating a brain is just a matter of simulating them all at once. The only caveat I know of to that is if there was somehow quantum interaction among multiple neurons that caused them to interact as if the (classical) whole were more than the sum of the (classical) parts, but for that to happen there would have to be quantum coherence on a scale of multiple neurons, and quantum coherence on that scale is very delicate and rare - the only examples I know of happen in a vacuum, barring superfluids, superconductivity, and bose-einstein condensate (which occur almost exclusively at temperatures near absolute zero).
I agree with you that something like this is the most likely way to make simulating a brain work, but...
There are very interesting ethical issues here. If you simulate my brain but with some errors that make it batshit nuts, or leave it writhing in agony, are you torturing someone? Does turning the program off make everything OK?
How many simulated people do you torture, then terminate while trying to get it right? Why should it matter if the consciousness of the critter tortured is based on physical interactions of atoms or on the flow of electrons?
I'm agnostic, and I *am* picking a side and standing up for it.
'Agnostic' doesn't mean "I don't know whether to believe in the Christian god" (as your comment re: Zeus appears to imply). Agnosticism can mean either "I don't know whether there's a god or not" or "I believe it's unknowable whether there's a god or not".
A god could take many, many forms. I believe there's strong evidence against a god who created humanity in its image and for the purpose of exalting it.
Not having heard any reasonable theory of the origin of the stuff in the universe, the space-time in which it sits, and the physical laws governing that stuff, I don't know where it comes from. I find it equally hard to believe that some entity outside of the bounds of physical laws created it or that it has no origin. (Giving a physical explanation of its origin just begs the question of the origin of the physics of the explanation).
I don't think you know either. So I call on you, stop pretending you're omniscient: admit you're not atheist, but rather agnostic.
Actually, I have a BS in Physics and Math. I said I had no special training in food science or metabolism, not that I'm an uneducated idiot.
Unless and until someone develops all these systems (starting with file transfer) to operate on challenge & response using pub/priv key systems. Hmm, in fact I suspect ssh does work that way when you use keys. I need to look into that...
Anyway, if you use such a system, then you can have a token (physical key) which contains your private key in an inaccessible way, but can provide the public key on demand or can be asked to sign a piece of text using your private key. Then the authenticating system sends you random text, your token signs it & sends it back, and the authenticating system validates the signature using your public key.
Then, regardless of the security of the system you're using, your identity is secure. Someone may steal your session and do Bad Things, but they can't ever login again as you once you're logged out on that session. For very dangerous operations, the authenticating system just requests you to authorize the operation explicitly, so even session hijacking won't let an attacker screw you. If necessary your token can provide a way to display a simple message about what operation you're authenticating by pressing OK.
Ideally, the physical key has a way for you to enter a password directly into it. Then someone would have to not only steal your key, but also learn your password to be able to hijack your identity.
If you need to invalidate an old token and migrate to a new one, you can allow a request signed by some authority to invalidate the old key and inject a new one. If you don't like central authorities, web of trust can be used in the same way.
Weight gain is *mass* in minus *mass* out. Now, how much of the mass of what you eat sticks with you may be related to the calories in it. In fact, it seems downright likely. But it is not the end-all.
This notion of "duh, OBVIOUSLY if you consume more calories than you burn, you'll gain weight" is pseudoscientific thinking. Calories are a measure of the chemical energy in your food. Whether your body can access that energy, how it stores it, how the food affects your metabolism, and a whole host of other things will affect weight gain.
I have no special training in food science or metabolism, but the "obvious" platitudes spouted off here wrt weight gain are obviously wrong.
I agree that he should carry through the "self-reported" modifier when he talks about the group with his preferred response, but the excerpt you gave doesn't have any problem in the science. He isn't saying the results are from people who are actually good at math - it clearly says "self-reported".
So in your view, there is no difference between "images of minors engaged in simulated sex acts" and "simulated images of minors engaged in sex acts"?
Because that's the difference the post is discussing, not whether you think either activity is OK.
Oh, duh, thanks.
That wasn't meant to be critical; that was honest wtf-ness.
Holy shit, your user id is huge! I didn't know we had made it into the tens of millions.
Cut-n-paste:
Factoring in all these issues, various groups have attempted to calculate a true economic cost for electricity generated by the most modern designs proposed. Because if an actual cost per kW*h can be calculated, then it is possible to compare it to other power sources to determine if such an investment is economically sound.
In 2003, the Massachusetts Institute of Technology (MIT) issued a report entitled, "The Future of Nuclear Power". They estimated that new nuclear power in the US would cost 6.7 cents per kW*h.[1] However, the Energy Policy Act of 2005 includes a tax credit that should reduce that cost slightly.
The lifetime cost of new generating capacity in the United States was estimated in 2006 by the U.S. government (the 2007 report did not estimate costs). Nuclear power was estimated at 5.93 cents per kW*h. However, the "total overnight cost" for new nuclear was assumed to be $1,984 per kWe - as seen above in Capital Costs, this figure is subject to debate.
A 2008 study based on historical outcomes in the U.S. said costs for nuclear power can be expected to run $0.25-.30 per kW*h.
A 2008 study concluded that if carbon capture and storage was required then nuclear power would be the cheapest source of electricity even at $4,038/kW in overnight capital cost.
In 2009, MIT updated its 2003 study, concluding that inflation and rising construction costs had increased the overnight cost of nuclear power plants to about $4,000/kWe, and thus increased the power cost to 8.4c/kW*h.
http://en.wikipedia.org/wiki/Economics_of_new_nuclear_power_plants#Cost_per_kW.C2.B7h
Interesting - the fellow I know is an extreme christian who actually avoids learning about quantum mechanics, because he doesn't want it to be true (?!), I'm not sure of his take on big bang, he believes in limited evolution (some speciation, but not that mammals and reptiles have a common ancestor, nor the animals having a common ancestor with man).
I know him from working on software dev projects with him, but he is one of my few work friends who are also friends outside of work. He and I are both National Merit Scholars; his background is in marine biology.
He was not a christian until probably his 30s - his wife "turned" him. Now he is a deacon of his church.
Oddly, he is not an idiot.
In case you're interested: http://en.wikipedia.org/wiki/Mass%E2%80%93energy_equivalence
I actually have a Neo 1973 - no need for the capacitor, nor is there a need for it in the newest Freerunners. I do agree that he should read up on it and hang out in #openmoko on irc.freenode.net to see what to expect. Early on, software was very buggy. Now, afaik, the only persistent problem is short battery life (about a day with normal usage).
"it hardly works" is inaccurate. There are issues he should understand *is* accurate.
BTW, the first line of http://en.wikipedia.org/wiki/Mass%E2%80%93energy_equivalence is:
In physics, mass-energy equivalence is the concept that the mass of a body is a measure of its energy content.
I'll try one more time. Please read http://en.wikipedia.org/wiki/Mass%E2%80%93energy_equivalence
I do actually have a Bachelor of Science in Physics and Mathematics, double major...
And, just to point out, an electromagnetic field can create positron/electron pairs, and both positrons and electrons have (positive) mass.
Get a Freerunner. It's an open phone with: GPS, internet via gprs, accelerometers, full bluetooth, wifi, 640x480 touchscreen. It runs any of several flavors of Linux (including Debian or Android, but my personal choice is SHR) and there are already Free gps programs that use OpenStreetMap (TangoGPS or Navit).
It's about $250, IIRC, but of course you can carry it around and use it to browse the web and receive calls, as well as using it in-vehicle for navigation.
It doesn't get any hackabler.
As the AC said, the chemistry doesn't matter. Energy is mass; it's just hard to convert significant amounts of one to the other. It doesn't matter if the energy is chemical, nuclear, kinetic, gravitational.
It's just (much) easier to see the change in mass in a nuclear reaction than in a chemical one.
BTW, if the released energy from the chemical reaction all went to heat in the solution, none of which escapes, then the mass of the solution wouldn't change. All that matters (no pun intended) is how much energy enters or leaves the system - there is a corresponding change in mass.