He basically said California's Voting Laws were so complex and constantly changing that they were not upset at having to leave the CA e-voting machine market.
Sounds like the pot calling the Kettle Black to me.
The problem isn't interface...its the inability of some people to understand how computers work. And pretty UI's don't fix that.
One purpose of a good interface is to abstract away from the details of how the computer works.
Some people have better things to do with their time then to learn how a computer works. Particularly if it has nothing to do with a job or hobby since they will promptly forget most of these details because of lack of use and reinforcement. I'm sure you have a TV in your house. Do you know how to fix it.
How come that web site is able to see printers on my computer? Is it a cookie leak like Internet Explorer?
That website is not on the Internet. Localhost is mapped to your own machine and the 631 is a port where CUPS is listening. A browser is a useful interface tool for local programs since it has nice, easy ways to display information.
Are you high right now? The problem with drugs isn't that they are
illegal to purchase or posess, it's that theyre DRUGS. Look how many
lives are devastated or lost by using/abusing legal pain medications,
or alcohol, or even tobacco.
And many people believe these problems would be worse with prohibition.
You can't have a perfect world, and prohibition doesn't get rid of
drugs. In many peoples eyes, prohibition just causes more problems
than it solves.
Saying there should be no laws against use/abuse of clearly harmful
substances is just wrong.
So we should outlaw cigarettes, alcohol, big macs, motorcycles,...
From the civil rights aspect, sure, let people get all
the drugs they want... it's their choice, it's their life.. right?
Well, what happens when they overdose? Leave them in the streets
because they dont have health insurance? Or do we hospitalize
them... give them medicine... rehab them... ??? With who's
money.. this would costs tens of thousands of dollars per person every
time they're found in the street?
Why would it cost so much? And legalized drugs would make it less
likely to overdose. Also legalizing drugs would most likely turn more
people on to lighter drugs for the same reasons hard liquor use went
down after prohibition. People don't overdose on weed. (They
overdose on alcohol all the time, but never weed.)
With my money? I think not! Because of society's general ideal of
helping those in need, we will never let some drug overdose "victim"
(or, no longer a victim, actually, so we'll say subject) lay in the
street dying. This is where the problem is.
So lets lock them up in jail at even greater cost. Also your
assuming we would have more overdoses with legalization. This is most
likely false.
You would not be able to successfully tax illicit drug sales in order
to defray the cost of hospital care, and even if you could, you'd be
selling it for more than it could be bought on the streets.
Most drugs are super cheap to make. It would be easy to get rid of
the blackmarket. When cocaine was legal, it was around the same price
as aspirin. Also any money generated is better than the super money
sink hole that currently exists. Who do you think is paying for this
war on drugs?
The second issue is the known crime caused by drug addicts. Because
of the addiciton caused by stuff like heroin, addicts will do
anything they can to obtain the drug. This includes theft,
prostitution, or in the worst cases assault and murder for hire. In
fact, a murder can occur simply because someone was attempted to get
$20 for his next high. This is a public safety issue in general.
This is an effect of prohibition. I don't hear about crime waves
for people to get their cigarettes which are extremely addictive.
The third issue is the quality of the work force. Legalizing drugs
which impair judgement would eventually mean companies could not
screen individuals (although, at first drug screening would hold up,
it wouldnt be long before the ACLU challenged the privacy legality of
the tests). The cause would be an ineffective work force, forcing
companies to go out of business. This could also cause hostilities in
a work place.
What makes you think most people in the work force (except you
presumably) is going to go to work high. Is everybody at your job
drunk? As a side issue, current drug testing is a farce. It just
encourages people to harder drugs since they are more difficult to
test. If you want to stop people from being high at work, fire them
if they are high, not if they were high two weekends ago.
The fourth point, is that sentences for selling to minors would be
just as lax as they are now with alcohol and cigarettes. Society's
view would be one of "well he's just acting more drown up" because
drug usage would be considered an "adult act".
Take about a straw man. Just as we can legalize drugs we can stiffen
up the penalties for selling to minors. Most likely we would have a
different system of distribution that didn't use neighborhood stores.
It's not fair to stipulate how the system of legalization will work
and then tear down your own bad ideas. Try tearing down someone elses
good ideas. There lots of information on the web. Do some research.
The fifth and final point, do we want more corporations like "big
tobacco" running our lives? Would they be required to state the
inevitable side effects of the usage of their product? Would they be
responsible for crimes being committed by those addicted to their drug?
Would they be civilly sued for not disclosing that their drug
PBS information
on
The Diamond Age
·
· Score: 5, Interesting
It looks like the big breakthrough is the CVD technique. The old
Russian design had the problem of letting in too much nitrogen and
creating only yellow diamonds. They have improved the technique but
it is still harder to make clear diamonds. I read that they were
going after the colored market since colored natural diamonds are more
expensive. Plus it must be easier to add color with new elements than
remove all the yellow. (They can add different elements to get
different colors.) Expect the market in colored diamonds (especially
yellow) to get cheap. (Kobe should have waited...) Of course the
real volume is in clear diamonds. Hopefully the CVD technique can
make cheap clear diamonds. I know they said $5 a carat, but I wouldn't trust Wired.
Members are all just trying to represent the voters and win
re-election.
I think the key phrase is win re-election. I'm not convinced that
representing the voters is the best way to get re-elected. More often
then not, the one who wins the the election is the one who spends the
most money. If this is true then the goal of the savvy politician is
to raise as much money as possible to ensure re-election. This is why
most people feel politicians are bought.
Of course, as you pointed out, because of the effect money has on
the process it is still possible to influence the system even if the
politicians are not corrupt. If an organization wants to influence
the laws they just fund the candidates with similar opinions. The more
candidates there are for a position, the more likely one will have a
platform that is similar. The problem is that this allows the people
with money to have an disproportionate influence on our laws.
The solution would seem to be campaign finance reform, however
meaningful campaign finance reform is unlikely because it makes it
harder for the existing politicians to get re-elected. And we are
back where we started...
That howstuffworks is pretty misleading. CD's are supposed to create sine waves not square waves. The output eventually goes through something that has the effect of a linear filter. This will smooth out the signal and give something close to a sine wave up to 20kHz. Some people think that the filter creates all kinds of audible artifacts and that's why modern CD players use oversampling to raise the frequency of the filter.
I disagree; it was low entropy. Entropy has been increasing ever since. Low entropy means the energy is concentrated in one place; high entropy means it is randomly distributed.
Just to be a trouble-maker, how do you define the size of the space. If the size of the space that holds the energy is changing then you can lower the entropy. How is the size of the universe defined?
I think you could simplify your argument to monopolies are bad.
It seems somewhat convoluted that the government would use the GPL to fight monopolies instead of the various anti-monopoly laws. Of course, the government isn't doing a good job applying these laws, but that's not because the laws don't have teeth, it's just that the government chooses not to bite the corporations.
He's taken money from the very people that his legislative plans
will benefit. Can any sane, rational person honestly believe that this
is not a conflict of interest? This is not right, and it's symptomatic
of the legalized bribery that is the core problem of the American
political system.
While I agree something smells fishy, other explanations are possible.
If there are enough people in office (or running for office), it
increases the odds that some of these people will share a corporations
views or can be persuaded to share these views. Once these people
agree with the corporation, naturally the corporation funds their
campaign.
The corporations you mentioned might have persuaded Wexler that the
he can get the most votes by pushing this agenda. The sticking point
is how he thinks he can get these votes. Does he think he can get the
votes from the great new policy, or does he think he will get the
votes by using the money for campaign advertisements on unrelated
issues.
Well, I'm not going to go through the effort of proving it, but as
I said log3(n) is probably a lower bound.
If you look closely at the technique I gave, when I determine if
the ball is heavy or light, I only use two value of the comparison,
not all three. This combined with some extra leverage when n!=3^k,
and you can probably squeeze out an extra comparison in some cases.
(If I had to guess, I'd say when n=3^k+b and b<=3^(k-1).)
You can do it in ceiling(log3(n))+1. On the first step divide the
marbles into three close to equal sized groups. (The groups may
differ by at most one marble.) Compare group 1 to group 2. If both
groups are equal, you know the strange marble is in group 3 and you
can use one more weighing to see if it is light or heavy. If they are
not equal then call group A the lighter group and group B the heavier
group. Compare group A to group 3. If these groups are equal then
the strange marble is in group B and it is heavy otherwise group A
must contain the strange marble and it is light.
So far we have used 2 comparisons. The rest is trivial. Just keep
on splitting into three close to equal sized groups. Each split will
remove about 1/3 of the marbles and therefore you get log3(n). The
ceiling comes from the fact that the number of marbles may not be a
power of 3, but it is not difficult to modify a rigorous proof for
n=3^k to the general case by adding the ceiling function.
(Intuitively, if n=3^(k-1)+b where b < 2*3^(k-1) then you can show that
assuming n=3^k must give an upperbound.) The plus one comes from the
extra comparison needed at the first step to determine if the strange
marble is heavy or light.
The key here is that the comparison model has three outcomes.
Whether group 1 equals group 2, group 1 is less than group 2, or group 1
is greater than group 2. Most likely one could prove, using arguments
similar to comparison sort proofs, that log3(n) is a lower bound.
The most common reason is that most performance hacks and
optimizations are not decidable, and you want a compiler to implement
only decidable algorithms because those are the ones that enable a
compiler to be deterministic. It is usually much easier for a person,
i.e., human, to determine what can be done, than it is for a machine
to determine that exact same thing.
Algorithms are deterministic so there is no way the compiler is going
to be nondeterministic. And I doubt that undecidable problems have
much to do with practical optimization. The efficiency of the
technique is the issue. Undecidability has to do with infinite sets
of problems. If an optimization was so great, I'm sure people would
be happy with a solution that only covered programs of a certain size.
The problem is that the optimization is way too expensive.
Does f always return true? Only if we can prove that a and b never
point to the same array. A person maybe able to do this, but a machine
would have great difficulty (assuming the machine could even do
it).
If a human can prove it so can a machine. What advantage to you
think humans have. The intuition behind your idea is that the
programmer himself should be able to realize if f always returns true
since he wrote the code. However the programmers task is much easier
than the compilers. The programmer knows how he wants the code to
behave. The compiler only knows how the code behaves. Lots of human
optimizations or missed optimizations turn out to be bugs.
Re:Complex numbers shouldn't make physical sense
on
Imagining Numbers
·
· Score: 1
Your statement about "abstract two dimensional objects" actually hints
that it's not hard to give complex numbers a real physical meaning to
explain them to someone...for exmaple, real numbers could be
longitude, and imaginary latitude, then a point on a map could be a
complex number.
If this was the case, one would just use a two dimensional vector.
Why bother with a complex number. The interesting thing about complex
numbers is how they work algebraically. If you want to give a complex
number a real physical meaning then you need to give a physical
interpretation to addition and multiplication.
Even then it's somewhat besides the point. The primary use of complex
numbers is as a mathematical tool to solve real valued problems. The
solutions we care about deal with real values. Complex numbers
numbers just make it easier to solve the equations.
Of course, this doesn't mean that complex numbers are uninteresting.
It's very interesting to look at the extra properties they give us
that make it easier to solve problems. However, I think it's the
reason people find them so confusing. People think there must be some
real world interpretation of a complex value, when in fact, the first
thing one does after solving the problem is throw away all the complex
solutions.
Complex numbers shouldn't make physical sense
on
Imagining Numbers
·
· Score: 1
The reason complex numbers are so hard to understand is because they
are rarely used to model the real world. Real numbers are intuitive
because they are generally used to represent a magnitude. The
variables in a problem often represent real numbers. However, for
some problems, it becomes very difficult to work with real numbers.
This is where complex numbers come to the rescue.
Complex numbers have extra properties that make it easier to solve
problems, and they are a superset of the real numbers. To solve a
problem, just assume the variables are complex and generate all the
solutions. Any real solution to the original real problem must be a
solution to the complex complex problem, and any real solution to the
complex problem must be a solution to the original real problem.
Therefore, you just need to generate all the solutions and throw away
any complex solutions.
This is how complex numbers are used in practice. They are just a
mathematical tool. Without this burden of giving complex numbers a
physical interpretation, (Though this is still possible for some types
of problems) it makes more sense to view them as abstract two
dimensional objects. Addition is just vector addition and
multiplication is scalar multiplication along with rotation.
This is one of the main ways math is generalized. By adding extra
properties to an object, it makes it easier to work with the object.
This can be seen in the historical changes in the concept of a number.
From natural, integer, rational, real, and complex. By adding more
structure, the object actually becomes easier to use.
Of course, the another way to generalize is to take a result and strip
away all the unnecessary details. For example, one starts with
calculus on intervals and then proceeds to metric spaces and then
topologies...
All belief systems are not the same. In science, one chooses
empirically justified axioms. As long as there is a common bond of
experience, we can model those experiences with axioms. We pick and
choose things from the infinite world of mathematical truth to
imperfectly model the real world. If our axioms are very close to
reality, we can derive many levels of real consequence.
Of course, you can pick any set of axioms and tie those axioms to
the world in strange and imprecise way. This lack of rigor makes it
impossible to derive consequences of any meaningful depth. Even if
you could derive new theorems, why should they help describe the
objects they model. If the axioms don't describe experience, why
should the theorems.
His original question seems interesting and reasonable to me. He is asking how much does it cost to produce an album, and his math is meant to show that the average album can't cost $1,000,000 because the record companies would be losing money.
In fact, some of the best albums cost very little to produce. (Advertisement is another story.) Using minimalist microphones in a real venue as opposed to a multitrack studio recording, can get some of the most realistic results. Just listen to the Cowboy Junkies Trinity sessions. These types of recordings can be made with $50,000 worth of equipment. This is to own, not rent the equipment.
Did it ever occur to you that mass might not be very effective at
distorting space, and that energy might be? Remember that we are
talking about local distortions - not large ones like a planet
causes.
In modern physics, mass and energy are equivalent.
Once again the key questions are: "Does it work? Is it
repeatable". If the answer to those questions are "Yes" nothing else
matters. Theory has to be adjusted to fit reality - not the other way
around.
Yes, someone has to come up with a hypothesis and test it with
experiments. This may not require any changes in the theoretical
foundations. It may just require a better understanding of how
current theory applies to the problem.
For example, one people noticed that the voyager probes were
decelerating more rapidly then predicted by theory, the first thing
the scientists did is try to explain the phenomena using the existing
theory. Only after these explanations have failed, would people
consider changing the laws of physics and spending money to test these
new laws. Fortunately for the lifter problem, it should be easier to
test theories since it is so easy to repeat experiments.
Re:We (probably) won't ever actually ACHIEVE AI
on
Arguing A.I.
·
· Score: 3, Insightful
The problem with AI is that it always seems unsuccsessful. Any time an
AI technology matures and becomes useful it is no longer considered
"AI". Computer vision (face recognition), expert systems, even many
modern strategy games would be considered amazing AI advances a few
decades ago. They all arose because of AI research. Once they matured,
however, they were no longer considered AI.
The reason it is unsuccessful is the confusion caused by the
different meanings of the phrase AI.
Often AI just means research on a specific problem that humans are
currently much better at solving than machines. Of course once the
research is complete and the machine is better, it is no longer AI
under this definition.
Now if the solution is largely motivated by what we know about how
humans work then perhaps there is still a glimmer of AI in the
research. However, this is a hard argument to make since we don't
know how the brain works. In fact, often there are many reasons to
think the solution isn't similar to the brain. There are many ways to
skin a cat. For example, I doubt human chess masters search a game
tree with alpha-beta pruning, however, this is a way for computers to
solve the problem that, with todays hardware, gives them superior
performance.
AI won't be considered successful until we build HAL or Data, but the
journey so far has been very useful.
This is a different notion of AI. It fits more into the natural
definition of AI, where AI is the creation of human intelligence. In
this case, you need the whole enchilada (or at least a interesting
percentage) to get intelligence. You can't just pick and choose
certain problems. This definition is more in line with the Turing
Test. Unfortunately this is a very hard problem for obvious reasons.
At one time more people worked on this problem, but when nobody got
good results, the funding started to dry up. That's why people
switched to the previous definition.
Some people still work on the grand AI problem, but as another
poster pointed out, it is generally on a small piece with a story
about how it can be connected to other pieces to create a real AI.
Generally they pick a piece that might be commercially useful in its
own right such as vision or linguistics. Again this helps with
funding. Unfortunately, I don't think anyone works on tying these
systems together. (Probably because there would be a whole mess of
problems if they tried.)
The main performance benefit of a RAID is in reducing the impact of seek time on overall throughput.
(In other words, you do seeks 1/N as often...
What are you comparing against, because I just don't see why this would be true. I would figure the seek time would be about the some while the bandwidth would increase by a factor of the number of drives. Therefore, the seek time would have a greater effect on performance.
Pre-industrial man has been quite capable of committing massacres,
outright genocide and causing large scale eco-disasters.
I'm not sure what you mean by pre-industrial, but I'm using technology
in a broad sense. For example, wheels, swords, guns...
My claim is that accessible and advanced technology can make things
worse. A couple of vague examples is not sufficient to counter this
claim. You need to show that technology can't make these events more
likely. As someone else pointed out giving a nuclear bomb to everyone
in the world is a very dangerous proposition. Giving everybody a club
isn't such a big deal.
The claim shouldn't be too controversial. Technology can give more
power to individuals. Some percentage of people will abuse that
power. These people could do a lot of damage. Of course, not all
technology has this effect. That's why the individual issues need to
be debated.
//Too lazy to log in
Yeah, it takes more effort to login then type this.
Slashdot Mirror
He basically said California's Voting Laws were so complex and constantly changing that they were not upset at having to leave the CA e-voting machine market.
Sounds like the pot calling the Kettle Black to me.
The expression is sour grapes.
Not necessarily. The greatest engineering breakthroughs maybe. But not the greatest intellectual breakthroughs.
Look at computer science for example. People never thought about the existance of a "general purpose computing machine" till Bertrand russell came by.
You seem to have forgotten Charles Babage.
The problem isn't interface...its the inability of some people to understand how computers work. And pretty UI's don't fix that.
One purpose of a good interface is to abstract away from the details of how the computer works.
Some people have better things to do with their time then to learn how a computer works. Particularly if it has nothing to do with a job or hobby since they will promptly forget most of these details because of lack of use and reinforcement. I'm sure you have a TV in your house. Do you know how to fix it.
How come that web site is able to see printers on my computer? Is it a cookie leak like Internet Explorer?
That website is not on the Internet. Localhost is mapped to your own machine and the 631 is a port where CUPS is listening. A browser is a useful interface tool for local programs since it has nice, easy ways to display information.
I doubt it's that hard to finish with an SUV going 35 mph. Races are considerably faster.
Are you high right now? The problem with drugs isn't that they are illegal to purchase or posess, it's that theyre DRUGS. Look how many lives are devastated or lost by using/abusing legal pain medications, or alcohol, or even tobacco.
And many people believe these problems would be worse with prohibition. You can't have a perfect world, and prohibition doesn't get rid of drugs. In many peoples eyes, prohibition just causes more problems than it solves.
Saying there should be no laws against use/abuse of clearly harmful substances is just wrong.
So we should outlaw cigarettes, alcohol, big macs, motorcycles, ...
Why would it cost so much? And legalized drugs would make it less likely to overdose. Also legalizing drugs would most likely turn more people on to lighter drugs for the same reasons hard liquor use went down after prohibition. People don't overdose on weed. (They overdose on alcohol all the time, but never weed.)
With my money? I think not! Because of society's general ideal of helping those in need, we will never let some drug overdose "victim" (or, no longer a victim, actually, so we'll say subject) lay in the street dying. This is where the problem is.
So lets lock them up in jail at even greater cost. Also your assuming we would have more overdoses with legalization. This is most likely false.
You would not be able to successfully tax illicit drug sales in order to defray the cost of hospital care, and even if you could, you'd be selling it for more than it could be bought on the streets.
Most drugs are super cheap to make. It would be easy to get rid of the blackmarket. When cocaine was legal, it was around the same price as aspirin. Also any money generated is better than the super money sink hole that currently exists. Who do you think is paying for this war on drugs?
The second issue is the known crime caused by drug addicts. Because of the addiciton caused by stuff like heroin, addicts will do anything they can to obtain the drug. This includes theft, prostitution, or in the worst cases assault and murder for hire. In fact, a murder can occur simply because someone was attempted to get $20 for his next high. This is a public safety issue in general.
This is an effect of prohibition. I don't hear about crime waves for people to get their cigarettes which are extremely addictive.
The third issue is the quality of the work force. Legalizing drugs which impair judgement would eventually mean companies could not screen individuals (although, at first drug screening would hold up, it wouldnt be long before the ACLU challenged the privacy legality of the tests). The cause would be an ineffective work force, forcing companies to go out of business. This could also cause hostilities in a work place.
What makes you think most people in the work force (except you presumably) is going to go to work high. Is everybody at your job drunk? As a side issue, current drug testing is a farce. It just encourages people to harder drugs since they are more difficult to test. If you want to stop people from being high at work, fire them if they are high, not if they were high two weekends ago.
The fourth point, is that sentences for selling to minors would be just as lax as they are now with alcohol and cigarettes. Society's view would be one of "well he's just acting more drown up" because drug usage would be considered an "adult act".
Take about a straw man. Just as we can legalize drugs we can stiffen up the penalties for selling to minors. Most likely we would have a different system of distribution that didn't use neighborhood stores. It's not fair to stipulate how the system of legalization will work and then tear down your own bad ideas. Try tearing down someone elses good ideas. There lots of information on the web. Do some research.
The fifth and final point, do we want more corporations like "big tobacco" running our lives? Would they be required to state the inevitable side effects of the usage of their product? Would they be responsible for crimes being committed by those addicted to their drug? Would they be civilly sued for not disclosing that their drug
PBS had a special on this back in 2000. http://www.pbs.org/wgbh/nova/diamond/
It looks like the big breakthrough is the CVD technique. The old Russian design had the problem of letting in too much nitrogen and creating only yellow diamonds. They have improved the technique but it is still harder to make clear diamonds. I read that they were going after the colored market since colored natural diamonds are more expensive. Plus it must be easier to add color with new elements than remove all the yellow. (They can add different elements to get different colors.) Expect the market in colored diamonds (especially yellow) to get cheap. (Kobe should have waited...) Of course the real volume is in clear diamonds. Hopefully the CVD technique can make cheap clear diamonds. I know they said $5 a carat, but I wouldn't trust Wired.
Members are all just trying to represent the voters and win re-election.
I think the key phrase is win re-election. I'm not convinced that representing the voters is the best way to get re-elected. More often then not, the one who wins the the election is the one who spends the most money. If this is true then the goal of the savvy politician is to raise as much money as possible to ensure re-election. This is why most people feel politicians are bought.
Of course, as you pointed out, because of the effect money has on the process it is still possible to influence the system even if the politicians are not corrupt. If an organization wants to influence the laws they just fund the candidates with similar opinions. The more candidates there are for a position, the more likely one will have a platform that is similar. The problem is that this allows the people with money to have an disproportionate influence on our laws.
The solution would seem to be campaign finance reform, however meaningful campaign finance reform is unlikely because it makes it harder for the existing politicians to get re-elected. And we are back where we started...
That howstuffworks is pretty misleading. CD's are supposed to create sine waves not square waves. The output eventually goes through something that has the effect of a linear filter. This will smooth out the signal and give something close to a sine wave up to 20kHz. Some people think that the filter creates all kinds of audible artifacts and that's why modern CD players use oversampling to raise the frequency of the filter.
I disagree; it was low entropy. Entropy has been increasing ever since. Low entropy means the energy is concentrated in one place; high entropy means it is randomly distributed.
Just to be a trouble-maker, how do you define the size of the space. If the size of the space that holds the energy is changing then you can lower the entropy. How is the size of the universe defined?
I think you could simplify your argument to monopolies are bad.
It seems somewhat convoluted that the government would use the GPL to fight monopolies instead of the various anti-monopoly laws. Of course, the government isn't doing a good job applying these laws, but that's not because the laws don't have teeth, it's just that the government chooses not to bite the corporations.
While I agree something smells fishy, other explanations are possible.
If there are enough people in office (or running for office), it increases the odds that some of these people will share a corporations views or can be persuaded to share these views. Once these people agree with the corporation, naturally the corporation funds their campaign.
The corporations you mentioned might have persuaded Wexler that the he can get the most votes by pushing this agenda. The sticking point is how he thinks he can get these votes. Does he think he can get the votes from the great new policy, or does he think he will get the votes by using the money for campaign advertisements on unrelated issues.
Yes, I missed that post. Slashdot is not really a great forum for discussion.
Well, I'm not going to go through the effort of proving it, but as I said log3(n) is probably a lower bound.
If you look closely at the technique I gave, when I determine if the ball is heavy or light, I only use two value of the comparison, not all three. This combined with some extra leverage when n!=3^k, and you can probably squeeze out an extra comparison in some cases. (If I had to guess, I'd say when n=3^k+b and b<=3^(k-1).)
Well since no one else answered.
You can do it in ceiling(log3(n))+1. On the first step divide the marbles into three close to equal sized groups. (The groups may differ by at most one marble.) Compare group 1 to group 2. If both groups are equal, you know the strange marble is in group 3 and you can use one more weighing to see if it is light or heavy. If they are not equal then call group A the lighter group and group B the heavier group. Compare group A to group 3. If these groups are equal then the strange marble is in group B and it is heavy otherwise group A must contain the strange marble and it is light.
So far we have used 2 comparisons. The rest is trivial. Just keep on splitting into three close to equal sized groups. Each split will remove about 1/3 of the marbles and therefore you get log3(n). The ceiling comes from the fact that the number of marbles may not be a power of 3, but it is not difficult to modify a rigorous proof for n=3^k to the general case by adding the ceiling function. (Intuitively, if n=3^(k-1)+b where b < 2*3^(k-1) then you can show that assuming n=3^k must give an upperbound.) The plus one comes from the extra comparison needed at the first step to determine if the strange marble is heavy or light.
The key here is that the comparison model has three outcomes. Whether group 1 equals group 2, group 1 is less than group 2, or group 1 is greater than group 2. Most likely one could prove, using arguments similar to comparison sort proofs, that log3(n) is a lower bound.
Algorithms are deterministic so there is no way the compiler is going to be nondeterministic. And I doubt that undecidable problems have much to do with practical optimization. The efficiency of the technique is the issue. Undecidability has to do with infinite sets of problems. If an optimization was so great, I'm sure people would be happy with a solution that only covered programs of a certain size. The problem is that the optimization is way too expensive.
Does f always return true? Only if we can prove that a and b never point to the same array. A person maybe able to do this, but a machine would have great difficulty (assuming the machine could even do it).If a human can prove it so can a machine. What advantage to you think humans have. The intuition behind your idea is that the programmer himself should be able to realize if f always returns true since he wrote the code. However the programmers task is much easier than the compilers. The programmer knows how he wants the code to behave. The compiler only knows how the code behaves. Lots of human optimizations or missed optimizations turn out to be bugs.
If this was the case, one would just use a two dimensional vector. Why bother with a complex number. The interesting thing about complex numbers is how they work algebraically. If you want to give a complex number a real physical meaning then you need to give a physical interpretation to addition and multiplication.
Even then it's somewhat besides the point. The primary use of complex numbers is as a mathematical tool to solve real valued problems. The solutions we care about deal with real values. Complex numbers numbers just make it easier to solve the equations.
Of course, this doesn't mean that complex numbers are uninteresting. It's very interesting to look at the extra properties they give us that make it easier to solve problems. However, I think it's the reason people find them so confusing. People think there must be some real world interpretation of a complex value, when in fact, the first thing one does after solving the problem is throw away all the complex solutions.
The reason complex numbers are so hard to understand is because they are rarely used to model the real world. Real numbers are intuitive because they are generally used to represent a magnitude. The variables in a problem often represent real numbers. However, for some problems, it becomes very difficult to work with real numbers. This is where complex numbers come to the rescue.
Complex numbers have extra properties that make it easier to solve problems, and they are a superset of the real numbers. To solve a problem, just assume the variables are complex and generate all the solutions. Any real solution to the original real problem must be a solution to the complex complex problem, and any real solution to the complex problem must be a solution to the original real problem. Therefore, you just need to generate all the solutions and throw away any complex solutions.
This is how complex numbers are used in practice. They are just a mathematical tool. Without this burden of giving complex numbers a physical interpretation, (Though this is still possible for some types of problems) it makes more sense to view them as abstract two dimensional objects. Addition is just vector addition and multiplication is scalar multiplication along with rotation.
This is one of the main ways math is generalized. By adding extra properties to an object, it makes it easier to work with the object. This can be seen in the historical changes in the concept of a number. From natural, integer, rational, real, and complex. By adding more structure, the object actually becomes easier to use.
Of course, the another way to generalize is to take a result and strip away all the unnecessary details. For example, one starts with calculus on intervals and then proceeds to metric spaces and then topologies...
All belief systems are not the same. In science, one chooses empirically justified axioms. As long as there is a common bond of experience, we can model those experiences with axioms. We pick and choose things from the infinite world of mathematical truth to imperfectly model the real world. If our axioms are very close to reality, we can derive many levels of real consequence.
Of course, you can pick any set of axioms and tie those axioms to the world in strange and imprecise way. This lack of rigor makes it impossible to derive consequences of any meaningful depth. Even if you could derive new theorems, why should they help describe the objects they model. If the axioms don't describe experience, why should the theorems.
His original question seems interesting and reasonable to me. He is asking how much does it cost to produce an album, and his math is meant to show that the average album can't cost $1,000,000 because the record companies would be losing money.
In fact, some of the best albums cost very little to produce. (Advertisement is another story.) Using minimalist microphones in a real venue as opposed to a multitrack studio recording, can get some of the most realistic results. Just listen to the Cowboy Junkies Trinity sessions. These types of recordings can be made with $50,000 worth of equipment. This is to own, not rent the equipment.
In modern physics, mass and energy are equivalent.
Once again the key questions are: "Does it work? Is it repeatable". If the answer to those questions are "Yes" nothing else matters. Theory has to be adjusted to fit reality - not the other way around.Yes, someone has to come up with a hypothesis and test it with experiments. This may not require any changes in the theoretical foundations. It may just require a better understanding of how current theory applies to the problem.
For example, one people noticed that the voyager probes were decelerating more rapidly then predicted by theory, the first thing the scientists did is try to explain the phenomena using the existing theory. Only after these explanations have failed, would people consider changing the laws of physics and spending money to test these new laws. Fortunately for the lifter problem, it should be easier to test theories since it is so easy to repeat experiments.
The reason it is unsuccessful is the confusion caused by the different meanings of the phrase AI.
Often AI just means research on a specific problem that humans are currently much better at solving than machines. Of course once the research is complete and the machine is better, it is no longer AI under this definition.
Now if the solution is largely motivated by what we know about how humans work then perhaps there is still a glimmer of AI in the research. However, this is a hard argument to make since we don't know how the brain works. In fact, often there are many reasons to think the solution isn't similar to the brain. There are many ways to skin a cat. For example, I doubt human chess masters search a game tree with alpha-beta pruning, however, this is a way for computers to solve the problem that, with todays hardware, gives them superior performance.
AI won't be considered successful until we build HAL or Data, but the journey so far has been very useful.This is a different notion of AI. It fits more into the natural definition of AI, where AI is the creation of human intelligence. In this case, you need the whole enchilada (or at least a interesting percentage) to get intelligence. You can't just pick and choose certain problems. This definition is more in line with the Turing Test. Unfortunately this is a very hard problem for obvious reasons. At one time more people worked on this problem, but when nobody got good results, the funding started to dry up. That's why people switched to the previous definition.
Some people still work on the grand AI problem, but as another poster pointed out, it is generally on a small piece with a story about how it can be connected to other pieces to create a real AI. Generally they pick a piece that might be commercially useful in its own right such as vision or linguistics. Again this helps with funding. Unfortunately, I don't think anyone works on tying these systems together. (Probably because there would be a whole mess of problems if they tried.)
The main performance benefit of a RAID is in reducing the impact of seek time on overall throughput.
(In other words, you do seeks 1/N as often...
What are you comparing against, because I just don't see why this would be true. I would figure the seek time would be about the some while the bandwidth would increase by a factor of the number of drives. Therefore, the seek time would have a greater effect on performance.
I'm not sure what you mean by pre-industrial, but I'm using technology in a broad sense. For example, wheels, swords, guns...
My claim is that accessible and advanced technology can make things worse. A couple of vague examples is not sufficient to counter this claim. You need to show that technology can't make these events more likely. As someone else pointed out giving a nuclear bomb to everyone in the world is a very dangerous proposition. Giving everybody a club isn't such a big deal.
The claim shouldn't be too controversial. Technology can give more power to individuals. Some percentage of people will abuse that power. These people could do a lot of damage. Of course, not all technology has this effect. That's why the individual issues need to be debated.
Yeah, it takes more effort to login then type this.