You could also make the argument that if a state does a great deal to accommodate a particular industry operating within its borders, it's reasonable to create a tax targeting that industry.
Another thing that should be brought up is that this is a.484% tax (reduced from 1.5% in 1998) on about 30% of Microsoft's revenue. The $747m figure is the total of the back taxes (and it must be noted that paying this won't break Microsoft's bottom line). I really don't see why paying this should be a huge deal.
... The only reason Washington is bitching is because they have a big budget shortfall....
Actually, Washington has been bitching about this for years (when they've had a surplus). Every year they bring it up MS threatens to leave and they back off. Personally, I think they should say "fine, pay us what you owe us and leave - but you'll never be permitted to sell your products in this state again." They've been extorting the state for years and it needs to stop.
Not to mention their is no goddamn way they would leave. The investment in their campus structure alone would not fly with Shareholders. This is a PUBLIC COMPANY and such threats are laughable, at best.
No chance in hell. There's no conceivable way they could find a buyer for their campus without selling at a massive loss. And there's no conceivable way they could tell their employees to take their kids out of Washington schools and tell them to move to some low-rent state like Nevada without having severe turnover in the process.
Up until now, I'm sure the state of Washington has been too afraid of losing Microsoft to lean on them. But I think now they may be desperate enough to do what they should have done all along, and make Microsoft pay what they owe to maintain the infrastructure that their employees use.
You're liable for all costs stemming from anything tortious you do even if some of those costs are attributable to a preexisting condition.
This makes sense if you look at this reasoning:
The security cleanup was a direct result of the crime.
Those systems have to be tested extensively to verify that the problems have been fixed.
If the break-in had not occurred, the fixes could have been done without much fuss.
The costs are attributable both to the preexisting security problems and the guilty party's actions, but it's difficult to determine exactly how much was attributable to each.
The easiest solution is to make the guilty party shoulder the entire cost to avoid short-changing the victim.
The thing is that when you're just looking at data points as they come in, you can't tell if a distribution is well-behaved. It's possible for the data to look like it's coming from a well-behaved distribution, but then you run into an outlier that's way too far out.
Those power-law distributions have cut-offs for the shape parameter for the higher order moments to exist. If the shape parameter is near the boundary, then I would imagine that there's a good chance that you could end up with a sample that looks like all of your analysis methods are working properly, but really comes from a distribution that infrequently produces severe outliers.
The lack of positive results in gravitational wave detection and the Higgs search reminds me of the Michelson-Morley experiment. Sure I know that we're only scraping the bottom/top of the possible ranges for these phenomena, but I wonder if we aren't just killing time until the next Einstein comes along to explain that there is no luminous aether.
Assumptions are okay so long as you only treat them as elements of long-term planning that will need to be revised periodically. I think that's the only safe way to view financial models.
But the financial engineers have committed the unforgivable sin of truly believing in their assumptions because they create a pleasant reality where you can bound risk into a little box. Reality is far less forgiving.
Furthermore, when you look at throughput for physical transportation, you have to factor in the impact of deadheads (moving an empty truck, plane, or messageless pigeon somewhere so it can pick up a load). In the case of carrier pigeons, they just fly toward home, so if you want to send another message you're going to have to carry them back to the source by some other means of transportation which likely takes much longer than the pigeons flight. Averaging in the return trip, the crummy network is going to win.
As an aside, the problem of minimizing the deadheads for a set of shipping orders between a variety of points is NP complete.
Intelligence is about making accurate approximations. If you try to calculate everything exactly you'll never get anything done.
Because of this, I do not believe you can remove the chance of error from intelligence. Furthermore, I think that emotions can be classified as states of predisposition toward avoiding certain classes of error based on situational priorities.
I do not believe that it's possible to make a human or greater level AI without the ability to shift the between these sorts of states based on situation. Whenever humans "solve" problems, we always just trade one kind of cost or error distribution for another sort of cost or error distribution that we view as more preferable. I do not see any reason to anticipate machines being able to bypass that.
I'm not terribly familiar with all of the details of corporate finance, so I'm guessing on some things here, but I'd have to agree. Long-term yields on AAA corporate bonds are around 5% right now. You could conceivably fund the project by selling bonds which pay out over the projected operational lifetime of the plant.
They claim that giving out duck calls differentiates them from other tour operators that instruct their passengers to quack. They'll have to provide evidence to support that claim. The court may or may not agree with them depending on the strength of that evidence.
Exactly. If someone comes to town and asks about touristy things to do, I might want to be able to tell them to go on the duck ride with the duck calls. If competitors imitate "the one with duck calls", then I have to tell them to go on the duck ride with duck calls, but not that other one.
If they've built up an association between duck calls and amphibious tours, isn't that roughly the same as building up an association between a logo or slogan and a product?
If everyone puffs their numbers it doesn't really matter much since advertisers are going to be looking at audience sizes relative to other shows.
If every network is in charge of its own ratings, then you'll have a situation where everyone is constantly looking for ways to maximize their own numbers compared to other networks. But if all of the networks back a single entity with one methodology, it's not going to cause that problem.
Classical logic has a direct correspondence with classical set theory. This means that if you're working in a formal system that can't be rigorously defined in terms of classical set theory, you have to accept some shades of gray between true and false.
It's clear that an objective reality that can be described this way exists. However, information theory shows us that there must be a minimum information content needed to describe it this way and this information content is greater than what will fit in your skull.
You haven't caught onto the big secret: analysis of historical data is ridiculously susceptible to confirmation bias. You can't pull hard conclusions out of it without cherry picking and you're capable of cherry picking without realizing it.
Supposing that a band is interesting enough to listen to, I would think that they would be interesting enough to put together some interesting supplementary material: art, liner notes, etc.
It gives bands an opportunity to have a more creative platform for presentation than just a cover image and a track list, which is all a digital album amounts to at present.
I have had this need when reverse engineering and debugging algorithms in software. There are magic numbers in the formulas and I have no idea what they mean.
Additionally, if something like this was rolled into a more generalized search algorithm, it could be used the other way around. Google could know, for example, that a paper with the number 58.44 a lot of times is probably about NaCl even if it is not mentioned explicitly.
Sorry, you're out of luck there. Magic numbers work because they're actually magical. Sorry to burst your bubble if you still believe in things like algorithms and the tooth fairy.
Did you honestly believe that computers worked by interpreting a series of logical instructions input by a mortal human being? Trust me. When you're ready to understand you won't need to search.
Chances are they want to provide things like DVD style menus that work consistently across a range of devices with different display and input capabilities. That's not a terrible idea and it is the sort of thing that you need some sort of standard for beyond "just use HTML".
I think this is a subtle point. The human race, as a whole, has a wide distribution of mental configurations and they all have their strengths and weaknesses. It's foolish to believe that you can just "design" something better without introducing commensurate faults. There are inherent trade-offs between, for example, intuitive thinking and rigorous logic. Both sides fail in different ways.
I have a suspicion that if you try to pack all of the strengths in one "brain" you're going to run into some of the hard physical constraints on information and computation. I think that the reason human error exists is that there's only so much information you can fit into a skull, and if you start trying bigger skulls then you're going to run into I/O limitations.
Slashdot Mirror
You could also make the argument that if a state does a great deal to accommodate a particular industry operating within its borders, it's reasonable to create a tax targeting that industry.
Another thing that should be brought up is that this is a .484% tax (reduced from 1.5% in 1998) on about 30% of Microsoft's revenue. The $747m figure is the total of the back taxes (and it must be noted that paying this won't break Microsoft's bottom line). I really don't see why paying this should be a huge deal.
Actually, Washington has been bitching about this for years (when they've had a surplus). Every year they bring it up MS threatens to leave and they back off. Personally, I think they should say "fine, pay us what you owe us and leave - but you'll never be permitted to sell your products in this state again." They've been extorting the state for years and it needs to stop.
Not to mention their is no goddamn way they would leave. The investment in their campus structure alone would not fly with Shareholders. This is a PUBLIC COMPANY and such threats are laughable, at best.
No chance in hell. There's no conceivable way they could find a buyer for their campus without selling at a massive loss. And there's no conceivable way they could tell their employees to take their kids out of Washington schools and tell them to move to some low-rent state like Nevada without having severe turnover in the process.
Up until now, I'm sure the state of Washington has been too afraid of losing Microsoft to lean on them. But I think now they may be desperate enough to do what they should have done all along, and make Microsoft pay what they owe to maintain the infrastructure that their employees use.
Eggshell skull rule
You're liable for all costs stemming from anything tortious you do even if some of those costs are attributable to a preexisting condition.
This makes sense if you look at this reasoning:
If you don't make distributional assumptions, there's little you can conclude about the nature of any distribution.
I guess really my issue is that I think distributional assumptions are dangerous, particularly when you're modeling the behavior of crowds.
These sorts of systems can shift behaviors if they run into a constraint that hasn't revealed itself yet.
The thing is that when you're just looking at data points as they come in, you can't tell if a distribution is well-behaved. It's possible for the data to look like it's coming from a well-behaved distribution, but then you run into an outlier that's way too far out.
Those power-law distributions have cut-offs for the shape parameter for the higher order moments to exist. If the shape parameter is near the boundary, then I would imagine that there's a good chance that you could end up with a sample that looks like all of your analysis methods are working properly, but really comes from a distribution that infrequently produces severe outliers.
The lack of positive results in gravitational wave detection and the Higgs search reminds me of the Michelson-Morley experiment. Sure I know that we're only scraping the bottom/top of the possible ranges for these phenomena, but I wonder if we aren't just killing time until the next Einstein comes along to explain that there is no luminous aether.
Assumptions are okay so long as you only treat them as elements of long-term planning that will need to be revised periodically. I think that's the only safe way to view financial models.
But the financial engineers have committed the unforgivable sin of truly believing in their assumptions because they create a pleasant reality where you can bound risk into a little box. Reality is far less forgiving.
Furthermore, when you look at throughput for physical transportation, you have to factor in the impact of deadheads (moving an empty truck, plane, or messageless pigeon somewhere so it can pick up a load). In the case of carrier pigeons, they just fly toward home, so if you want to send another message you're going to have to carry them back to the source by some other means of transportation which likely takes much longer than the pigeons flight. Averaging in the return trip, the crummy network is going to win.
As an aside, the problem of minimizing the deadheads for a set of shipping orders between a variety of points is NP complete.
Intelligence is about making accurate approximations. If you try to calculate everything exactly you'll never get anything done.
Because of this, I do not believe you can remove the chance of error from intelligence. Furthermore, I think that emotions can be classified as states of predisposition toward avoiding certain classes of error based on situational priorities.
I do not believe that it's possible to make a human or greater level AI without the ability to shift the between these sorts of states based on situation. Whenever humans "solve" problems, we always just trade one kind of cost or error distribution for another sort of cost or error distribution that we view as more preferable. I do not see any reason to anticipate machines being able to bypass that.
I'm not terribly familiar with all of the details of corporate finance, so I'm guessing on some things here, but I'd have to agree. Long-term yields on AAA corporate bonds are around 5% right now. You could conceivably fund the project by selling bonds which pay out over the projected operational lifetime of the plant.
The court can decide that.
They claim that giving out duck calls differentiates them from other tour operators that instruct their passengers to quack. They'll have to provide evidence to support that claim. The court may or may not agree with them depending on the strength of that evidence.
Exactly. If someone comes to town and asks about touristy things to do, I might want to be able to tell them to go on the duck ride with the duck calls. If competitors imitate "the one with duck calls", then I have to tell them to go on the duck ride with duck calls, but not that other one.
Why isn't this about brand identity?
If they've built up an association between duck calls and amphibious tours, isn't that roughly the same as building up an association between a logo or slogan and a product?
If everyone puffs their numbers it doesn't really matter much since advertisers are going to be looking at audience sizes relative to other shows.
If every network is in charge of its own ratings, then you'll have a situation where everyone is constantly looking for ways to maximize their own numbers compared to other networks. But if all of the networks back a single entity with one methodology, it's not going to cause that problem.
Thor doesn't damn. He smites.
So I think the word you're looking for is "thorsmitten".
Project Orion was certainly designed for planetary launch. They even did an analysis of how many people it would kill per launch due to fallout.
That's so metal.
I propose that we call this new constant 2pi.
Classical logic has a direct correspondence with classical set theory. This means that if you're working in a formal system that can't be rigorously defined in terms of classical set theory, you have to accept some shades of gray between true and false.
It's clear that an objective reality that can be described this way exists. However, information theory shows us that there must be a minimum information content needed to describe it this way and this information content is greater than what will fit in your skull.
You haven't caught onto the big secret: analysis of historical data is ridiculously susceptible to confirmation bias. You can't pull hard conclusions out of it without cherry picking and you're capable of cherry picking without realizing it.
There are still some semantics to specify and tyey
Supposing that a band is interesting enough to listen to, I would think that they would be interesting enough to put together some interesting supplementary material: art, liner notes, etc.
It gives bands an opportunity to have a more creative platform for presentation than just a cover image and a track list, which is all a digital album amounts to at present.
I have had this need when reverse engineering and debugging algorithms in software. There are magic numbers in the formulas and I have no idea what they mean.
Additionally, if something like this was rolled into a more generalized search algorithm, it could be used the other way around. Google could know, for example, that a paper with the number 58.44 a lot of times is probably about NaCl even if it is not mentioned explicitly.
Sorry, you're out of luck there. Magic numbers work because they're actually magical. Sorry to burst your bubble if you still believe in things like algorithms and the tooth fairy.
Did you honestly believe that computers worked by interpreting a series of logical instructions input by a mortal human being? Trust me. When you're ready to understand you won't need to search.
Chances are they want to provide things like DVD style menus that work consistently across a range of devices with different display and input capabilities. That's not a terrible idea and it is the sort of thing that you need some sort of standard for beyond "just use HTML".
I think this is a subtle point. The human race, as a whole, has a wide distribution of mental configurations and they all have their strengths and weaknesses. It's foolish to believe that you can just "design" something better without introducing commensurate faults. There are inherent trade-offs between, for example, intuitive thinking and rigorous logic. Both sides fail in different ways.
I have a suspicion that if you try to pack all of the strengths in one "brain" you're going to run into some of the hard physical constraints on information and computation. I think that the reason human error exists is that there's only so much information you can fit into a skull, and if you start trying bigger skulls then you're going to run into I/O limitations.
As an aside, how do you come up with a measure like "150 times colder"? I mean, what metric do you use to determine 1 times colder or 2?
Think in Kelvins.
So you're saying England is more like A Clockwork Orange than 1984?