Generalists are preferable because there is less communications overhead for one person than for a team. A generalist can move at the speed of thought, specialists move at the speed of staff meetings.
The infrastructure is fully depreciated and has already paid for itself several times over. ILECs don't like building out new plant anyway - they have no reason to since they have no competitors worthy of the name, they can't really charge more since their rates are already based on what the market can afford rather than the costs of providing service, and the new service would just cut into the sales or margins of their existing services. If you want the ILECs to compete, then you have to give them competition. If you want them to provide the best service technically possible at the best prices, then you have somehow instill a service culture like that of Korea's providers.
In the southeast US, the competing ISPs sold DSL service where everything from the local loop to the ISP connection was provided by the wholesale part of BellSouth. BellSouth's captive ISP was theoretically competing on the same basis as the other ISPs, but the volume clauses for price breaks were adjusted so that only BellSouth's ISP got much lower prices, and there was a huge amount of cooperation between the wholesale side of B.S. and the ostensibly independent ISP. The big pipes at the ISP end were also much cheaper for BellSouth's ISP. From the beginning B.S. had ~95% of the DSL market. I doubt that the situation was too different anywhere in the US - the ILECs all find ways to keep competitors off their lines through cost exaggeration, anti-competitive collusion and discrimination and regulatory capture.
The replies you got are ridiculously naive about how ILECs rig their accounting and prices. Audits don't help because the way that the accounting is rigged is actually legal. The "regulated" part of "regulated monopoly" is also pretty much a joke.
Beyond the government subsidies, the direct taxes to consumers, and letting the ILECs charge essentially whatever they wanted, the most important corrupt crony advantage the ILECs got was that they were granted a monopoly on essential services,a whole sector of the economy, and they got private title to all the publicly financed plant and equipment (which their captive subsidiaries made and marked up to ridiculous levels to launder the ILECs real profits).
I doubt your mom had any idea what it cost Verizon to actually provide the lines (it's a very fuzzy number, if you asked different accountants they could come up with virtually any number you wanted), likely all she had were the tariffs which are basically the equivalent of the sticker price on a car. Obviously the competitors were getting the lines much cheaper than the list price, and higher-ups at Verizon were either willingly selling them at that price or were unable to document higher costs to the regulators, even though ILECs inflate those cost claims to the limits of the abilities of their accountants.
Where the shenanigans happen is in the tariffs and in procurement. The rates for data lines do not drop after the line and the equipment have been paid off, nor do they fall along with the cost of the equipment, nor the actual costs of maintenance. Theoretically the state PUCs keep an eye on this, but in effect the ILECs charge whatever they feel like. Often the actual costs to the ILEC will be exactly the same for services that span one or even two orders of magnitude of price for the customer -the only difference between a 256M DSL and an 8M DSL is a few automatically set configuration entries in the DSLAM. The volume breakpoints for discounts on lines are set so that the ILECs ISP subsidiary is the only one who gets the really low rates. Similar but more obscure things happen with interdepartmental accounting for costs. An external customer will have to pay for every little thing, internal or affiliated customers don't. The tariffs are also set up to be as opaque and difficult to use as possible, thus deterring service requests for products the ILEC does not want to sell. Then the ILEC-written tariffs are gestured at as if they were law by the ILEC employees, but are virtually never read and often twisted to mean whatever is convenient. For instance "best-effort service" is typically taken to mean "we'll fix at it, maybe"
Procurement is another way of ripping off the customer - the ILEC overpays certain vendors, ones in which it either has a financial stake, (e.g. Western Electric) or, one suspects, ones which pay kickbacks in one form or another (perhaps Alcatel, Lucent). The ILEC can then say that DSL costs $1000/ per port or whatever, despite the true cost being much less. Together with treating the outside plant that has already been paid for three times over and fully depreciated on its taxes as being worth the same rent as when it was new, the ILEC can exploit its customers and push any competitors out of business. Repair metrics get twisted to be the number of calls or dispatches handled rather than whether issues were resolved to the customer's satisfaction, then the number of calls handled is increased by setting things up so that things take several calls to some third-world country to fix, if they get fixed at all. Then just to make things clear about ILEC priorities and the falsehood of their expense claims, field techs are given grossly inadequate test gear, when the appropriate equipment would only cost a week or two of the technician's customer billings (but they can bill more when things take longer, and bill more often if things never get fixed properly) .
I'm not sure either - the center of gravity looks a little high, but the same is true of a Segway - or a bicycle, for that matter - gyro-controlled adjustments of the thrust vectoring might be the answer, or this thing might have similar intrinsic stability to a bike, allowing manual thrust vector control plus rider weight shifting to control roll at speed. The stability in hover would still be a potential problem, even so. With a pitch-controlled prop capable of reversing airflow, together with the right kind of seatbelt, this could potentially fly inverted.
If the props were counter-rotating, then slight differentials of the relative rotational rate would allow yaw, but would have to be combined with prop pitch adjustments to maintain level attitude. It looks like there are small thrusters to handle this instead.
Fundamentally, I'm not convinced this thing has enough power to get out of ground effect, and I doubt the prop area is sufficient, and even if I'm wrong about that, the controls and stability are almost certainly going to be very tricky.
If they could prevent a bad script or plugin from taking the whole browser with it, that would be great.
It really needs a way to see a list of what scripts and plugins are running, what resources each is consuming and the ability to kill them individually. A list of currently open tabs would also be good, especially if it also had a history list for each tab (bonus points for making it editable, with drag and drop, etc.).
Even Google doesn't believe that it doesn't matter where the data is. When Kazakhstan said all.kz domains had to be hosted in that country, Google just walked away from providing Kazakh-tailored search. "If we were to operate google.kz only via servers located inside Kazakhstan, we would be helping to create a fractured Internet," said Google senior vice president for engineering Bill Coughran.
I hesitate to agree with Gartner about anything, but you can't trust that Google won't not only turn over your data to any jurisdiction that asks, but will likely cooperate with and not try to contest virtually any sort of court order or even law enforcement request. With a government-agency level of threat model, though, you shouldn't be storing information on computers that are ever hooked to the internet.
Sure. But you "don't want to sell beans in your store" because the city council told you not to - they want to reduce politically-oriented flatulence in town meetings.
But I won't buy anything from you at all, and I'm going to tell everyone you hate the tasty beans of freedom.
No, if T.F. were your doctor he'd insist that all cigarettes should be made in China, the US should devote itself to selling cigarette futures to each other, that you should have called the HMO's India-based "nurse" "helpline" before seeing him and that anybody smoking Turkish tobacco should be bombed and/or waterboarded. These insights would have come to him in a flash of some profoundly mixed metaphor that he would insist on boring you with at length before blaming it all on anti-globalist Arabs and all others less enlightened than Dr. F. (i.e. everybody else).
I agree with you, but I don't think this was what noted idiot, elite toady and NYT columnist (redundant, I know) Thomas Friedman had in mind. (Honestly, WTF? Why is that jackass' garbled, moronic gibberish showing up on Slashdot?)
Commodities are going up because the dollar is going down, and every other currency is following it so that their exports aren't at a disadvantage. The dollar is going down because the bank bailout never stopped, the Fed is funneling trillions to their cronies who then buy treasury bonds and thus keeping the interest rates down (thus making a risk-free profit, dramatically lowering the government's interest payments on the US debt, inflating the currency and incidentally making it hard for anybody else to get a loan for anything productive.)
The waste from photovoltaic production is quite nasty, and the composites for windmills are only better by comparison. The cost of and environmental damage to land by renewables is far greater than other sources of energy. Practical power storage technologies are not there yet, although they're coming along.
Many safer designs for nuclear plants are available, but even without them, nuclear is by far the safest type of power, far safer than PV or windmills, and over 10,000 times safer than coal. Nuclear energy produces less radioactive waste by any measure than coal, and high-level waste can be transmuted into short half-life isotopes as soon as somebody decides to build a subcritical reactor to do it. (As for heros, nobody is working in 4Sv/hr environments at Fukushima - that reading was taken by a robot. The exposure limits have rarely been reached by the workers, and even the new, higher limits are far below what has actually been shown to be harmful. The folks near the chemical plants and refineries that caught fire should be far more worried.)
Nuclear resources are not limited, there is easily enough thorium, depleted and natural uranium and "waste" out there for at least 10,000 years, and long before then we'll have figured out one of the other attractive energy possibilities. We'll run short of gallium, silver and other rare materials for PV long before that.
People who are anti-nuclear are anti-math, pro-poverty, and pro-death.
Perhaps some types of uranium breeders would work, although the sodium coolant in some designs and possibility of excursions in the reaction when burning weird actinide mixtures worries me a bit. I like the liquid thorium salt reactor the best of the breeders out there - safe, simple and complete fuel burnup, and it's already successfully tested.
A potentially better solution for secondary waste-burnup to supplement existing reactors would be subcritical reactors, which include accelerator-driven systems and hybrid fusion reactors. These generate neutrons at an energy cost using an accelerator or a sub-break-even fusion reactor to transmute wastes and generate net power from the wastes and from fertile materials such as thorium and depleted uranium. These reactors are potentially safer than the other fast reactors since they can instantly shut off the neutrons that sustain the reaction.
The health risks of radioactive isotope releases are overstated even for worst-case situations such as Fukushima. The answer to the deficiencies of old designs is not to give up what is by far the most ecologically responsible and lowest health risk way to generate power, but rather to build new nuclear plants using new technology.
Accelerator-driven and fusion-driven means of generating neutrons (at an energy cost) to irradiate fertile materials such as thorium and depleted uranium (providing a net energy gain) is a good example - these reactors can be shut off at the flip of a switch and can also generate power by transmuting high-level waste to forms that have much shorter half-lives, reducing its required storage to just decades - no long-term repository needed. Only a few of these are needed to handle the products of or provide the start-up isotopes for a much larger number of other advanced reactor designs, such as the nuclear candle, which can run sealed for decades, or the liquid thorium salt reactor, a proven technology which achieves virtually complete burn-up of its fuel and is intrinsically safe.
What would primarily determine the utility of Watson as a learning tool and its diagnoses as well, is the quality of the questions that Watson asks. Asking questions verifies the significance of the information it already has been given, helps distinguish between different possibilities, and leverages the capabilities of doctors by suggesting things too look at that they might not have considered. Watson is incapable of noticing anything; it depends on humans for that. Having the computer pointing out things for humans to look for uses the capabilities of both more effectively.
The feedback loop also needs to be closed - Watson needs to be told what worked and what didn't. I'd turn it loose analyzing the hundreds of millions of electronic patient records already in existence; there is far more information to be gleaned there than in medical journals.
I think it rearranged the wavefront a bit, which could be seen as a "partial collapse". There may have also been some times when certain photons were localized to one slit, but that would have been experimental noise rather than what they were trying to detect.
Short answer: who says ability is less heritable in Sweden? And even if so, who says that Sweden allows higher ability people to keep their earnings, or even tolerates the idea that some people just have more general ability than others?
Long answer: I'm not sure that the correlation is actually much less, rather the range of incomes is more restricted due to taxation and culture (not rewarding greater ability, denying its existence, or even ostracizing high ability people for insufficient humility). Also the Swedes are a more homogenous group than Americans, so the outcomes would also be expected to be more homogenous. OTOH, counter-intuitively, I believe the slighter differences might in some cultures actually lead to higher heritability of incomes (so long as we normalize to within-group z-scores), since it would take less of a shift in ability to change the outcome. But if the distribution of incomes is artificially narrowed relative to the distribution of abilities, as in Sweden, the reverse should happen. In America, with a winner-take-all mindset, of course you will see a wider spread of incomes, even after taking into account the more diverse population. And, as I said before, heritability isn't simply genetics - environments are also heritable, and they are also more diverse in America.
Yup, a good algorithm and strong key length are indispensable, but that's maybe 5% at most of what is needed for security. The algorithm per se is rarely the weak link - it's the implementation (there are always more wrinkles than anybody can foresee when writing such code), or the client computer with no physical security, a gazillion processes with more access than is safe, broken antivirus and firewall, insecure OS, poor configuration, browser vulnerabilities etc. ad nauseam.
Dropping SHA-bignum into an existing web application is like putting a bank-vault door into a camping tent.
Slashdot Mirror
Generalists are preferable because there is less communications overhead for one person than for a team. A generalist can move at the speed of thought, specialists move at the speed of staff meetings.
The infrastructure is fully depreciated and has already paid for itself several times over. ILECs don't like building out new plant anyway - they have no reason to since they have no competitors worthy of the name, they can't really charge more since their rates are already based on what the market can afford rather than the costs of providing service, and the new service would just cut into the sales or margins of their existing services. If you want the ILECs to compete, then you have to give them competition. If you want them to provide the best service technically possible at the best prices, then you have somehow instill a service culture like that of Korea's providers.
In the southeast US, the competing ISPs sold DSL service where everything from the local loop to the ISP connection was provided by the wholesale part of BellSouth. BellSouth's captive ISP was theoretically competing on the same basis as the other ISPs, but the volume clauses for price breaks were adjusted so that only BellSouth's ISP got much lower prices, and there was a huge amount of cooperation between the wholesale side of B.S. and the ostensibly independent ISP. The big pipes at the ISP end were also much cheaper for BellSouth's ISP. From the beginning B.S. had ~95% of the DSL market. I doubt that the situation was too different anywhere in the US - the ILECs all find ways to keep competitors off their lines through cost exaggeration, anti-competitive collusion and discrimination and regulatory capture.
The replies you got are ridiculously naive about how ILECs rig their accounting and prices. Audits don't help because the way that the accounting is rigged is actually legal. The "regulated" part of "regulated monopoly" is also pretty much a joke.
Nah, Microsoft's monopoly was a government monopoly, too - copyright.
Beyond the government subsidies, the direct taxes to consumers, and letting the ILECs charge essentially whatever they wanted, the most important corrupt crony advantage the ILECs got was that they were granted a monopoly on essential services,a whole sector of the economy, and they got private title to all the publicly financed plant and equipment (which their captive subsidiaries made and marked up to ridiculous levels to launder the ILECs real profits).
I doubt your mom had any idea what it cost Verizon to actually provide the lines (it's a very fuzzy number, if you asked different accountants they could come up with virtually any number you wanted), likely all she had were the tariffs which are basically the equivalent of the sticker price on a car. Obviously the competitors were getting the lines much cheaper than the list price, and higher-ups at Verizon were either willingly selling them at that price or were unable to document higher costs to the regulators, even though ILECs inflate those cost claims to the limits of the abilities of their accountants.
Where the shenanigans happen is in the tariffs and in procurement. The rates for data lines do not drop after the line and the equipment have been paid off, nor do they fall along with the cost of the equipment, nor the actual costs of maintenance. Theoretically the state PUCs keep an eye on this, but in effect the ILECs charge whatever they feel like. Often the actual costs to the ILEC will be exactly the same for services that span one or even two orders of magnitude of price for the customer -the only difference between a 256M DSL and an 8M DSL is a few automatically set configuration entries in the DSLAM. The volume breakpoints for discounts on lines are set so that the ILECs ISP subsidiary is the only one who gets the really low rates. Similar but more obscure things happen with interdepartmental accounting for costs. An external customer will have to pay for every little thing, internal or affiliated customers don't. The tariffs are also set up to be as opaque and difficult to use as possible, thus deterring service requests for products the ILEC does not want to sell. Then the ILEC-written tariffs are gestured at as if they were law by the ILEC employees, but are virtually never read and often twisted to mean whatever is convenient. For instance "best-effort service" is typically taken to mean "we'll fix at it, maybe"
Procurement is another way of ripping off the customer - the ILEC overpays certain vendors, ones in which it either has a financial stake, (e.g. Western Electric) or, one suspects, ones which pay kickbacks in one form or another (perhaps Alcatel, Lucent). The ILEC can then say that DSL costs $1000/ per port or whatever, despite the true cost being much less. Together with treating the outside plant that has already been paid for three times over and fully depreciated on its taxes as being worth the same rent as when it was new, the ILEC can exploit its customers and push any competitors out of business. Repair metrics get twisted to be the number of calls or dispatches handled rather than whether issues were resolved to the customer's satisfaction, then the number of calls handled is increased by setting things up so that things take several calls to some third-world country to fix, if they get fixed at all. Then just to make things clear about ILEC priorities and the falsehood of their expense claims, field techs are given grossly inadequate test gear, when the appropriate equipment would only cost a week or two of the technician's customer billings (but they can bill more when things take longer, and bill more often if things never get fixed properly) .
I'm not sure either - the center of gravity looks a little high, but the same is true of a Segway - or a bicycle, for that matter - gyro-controlled adjustments of the thrust vectoring might be the answer, or this thing might have similar intrinsic stability to a bike, allowing manual thrust vector control plus rider weight shifting to control roll at speed. The stability in hover would still be a potential problem, even so. With a pitch-controlled prop capable of reversing airflow, together with the right kind of seatbelt, this could potentially fly inverted.
If the props were counter-rotating, then slight differentials of the relative rotational rate would allow yaw, but would have to be combined with prop pitch adjustments to maintain level attitude. It looks like there are small thrusters to handle this instead.
Fundamentally, I'm not convinced this thing has enough power to get out of ground effect, and I doubt the prop area is sufficient, and even if I'm wrong about that, the controls and stability are almost certainly going to be very tricky.
If they could prevent a bad script or plugin from taking the whole browser with it, that would be great.
It really needs a way to see a list of what scripts and plugins are running, what resources each is consuming and the ability to kill them individually.
A list of currently open tabs would also be good, especially if it also had a history list for each tab (bonus points for making it editable, with drag and drop, etc.).
Even Google doesn't believe that it doesn't matter where the data is. When Kazakhstan said all .kz domains had to be hosted in that country, Google just walked away from providing Kazakh-tailored search. "If we were to operate google.kz only via servers located inside Kazakhstan, we would be helping to create a fractured Internet," said Google senior vice president for engineering Bill Coughran.
I hesitate to agree with Gartner about anything, but you can't trust that Google won't not only turn over your data to any jurisdiction that asks, but will likely cooperate with and not try to contest virtually any sort of court order or even law enforcement request. With a government-agency level of threat model, though, you shouldn't be storing information on computers that are ever hooked to the internet.
Sure. But you "don't want to sell beans in your store" because the city council told you not to - they want to reduce politically-oriented flatulence in town meetings.
But I won't buy anything from you at all, and I'm going to tell everyone you hate the tasty beans of freedom.
Jared Diamond is as big a weasel as Thomas Friedman (albeit weighing only about half as much).
No, if T.F. were your doctor he'd insist that all cigarettes should be made in China, the US should devote itself to selling cigarette futures to each other, that you should have called the HMO's India-based "nurse" "helpline" before seeing him and that anybody smoking Turkish tobacco should be bombed and/or waterboarded. These insights would have come to him in a flash of some profoundly mixed metaphor that he would insist on boring you with at length before blaming it all on anti-globalist Arabs and all others less enlightened than Dr. F. (i.e. everybody else).
And that's his good side.
I agree with you, but I don't think this was what noted idiot, elite toady and NYT columnist (redundant, I know) Thomas Friedman had in mind. (Honestly, WTF? Why is that jackass' garbled, moronic gibberish showing up on Slashdot?)
Commodities are going up because the dollar is going down, and every other currency is following it so that their exports aren't at a disadvantage. The dollar is going down because the bank bailout never stopped, the Fed is funneling trillions to their cronies who then buy treasury bonds and thus keeping the interest rates down (thus making a risk-free profit, dramatically lowering the government's interest payments on the US debt, inflating the currency and incidentally making it hard for anybody else to get a loan for anything productive.)
The waste from photovoltaic production is quite nasty, and the composites for windmills are only better by comparison. The cost of and environmental damage to land by renewables is far greater than other sources of energy. Practical power storage technologies are not there yet, although they're coming along.
Many safer designs for nuclear plants are available, but even without them, nuclear is by far the safest type of power, far safer than PV or windmills, and over 10,000 times safer than coal. Nuclear energy produces less radioactive waste by any measure than coal, and high-level waste can be transmuted into short half-life isotopes as soon as somebody decides to build a subcritical reactor to do it. (As for heros, nobody is working in 4Sv/hr environments at Fukushima - that reading was taken by a robot. The exposure limits have rarely been reached by the workers, and even the new, higher limits are far below what has actually been shown to be harmful. The folks near the chemical plants and refineries that caught fire should be far more worried.)
Nuclear resources are not limited, there is easily enough thorium, depleted and natural uranium and "waste" out there for at least 10,000 years, and long before then we'll have figured out one of the other attractive energy possibilities. We'll run short of gallium, silver and other rare materials for PV long before that.
People who are anti-nuclear are anti-math, pro-poverty, and pro-death.
Um, the linked numbers include pollution and show nuclear to be by far the safest.
"...polls consistently show that..." [citation needed]
Perhaps some types of uranium breeders would work, although the sodium coolant in some designs and possibility of excursions in the reaction when burning weird actinide mixtures worries me a bit. I like the liquid thorium salt reactor the best of the breeders out there - safe, simple and complete fuel burnup, and it's already successfully tested.
A potentially better solution for secondary waste-burnup to supplement existing reactors would be subcritical reactors, which include accelerator-driven systems and hybrid fusion reactors. These generate neutrons at an energy cost using an accelerator or a sub-break-even fusion reactor to transmute wastes and generate net power from the wastes and from fertile materials such as thorium and depleted uranium. These reactors are potentially safer than the other fast reactors since they can instantly shut off the neutrons that sustain the reaction.
The health risks of radioactive isotope releases are overstated even for worst-case situations such as Fukushima. The answer to the deficiencies of old designs is not to give up what is by far the most ecologically responsible and lowest health risk way to generate power, but rather to build new nuclear plants using new technology.
Accelerator-driven and fusion-driven means of generating neutrons (at an energy cost) to irradiate fertile materials such as thorium and depleted uranium (providing a net energy gain) is a good example - these reactors can be shut off at the flip of a switch and can also generate power by transmuting high-level waste to forms that have much shorter half-lives, reducing its required storage to just decades - no long-term repository needed. Only a few of these are needed to handle the products of or provide the start-up isotopes for a much larger number of other advanced reactor designs, such as the nuclear candle, which can run sealed for decades, or the liquid thorium salt reactor, a proven technology which achieves virtually complete burn-up of its fuel and is intrinsically safe.
What would primarily determine the utility of Watson as a learning tool and its diagnoses as well, is the quality of the questions that Watson asks. Asking questions verifies the significance of the information it already has been given, helps distinguish between different possibilities, and leverages the capabilities of doctors by suggesting things too look at that they might not have considered. Watson is incapable of noticing anything; it depends on humans for that. Having the computer pointing out things for humans to look for uses the capabilities of both more effectively.
The feedback loop also needs to be closed - Watson needs to be told what worked and what didn't. I'd turn it loose analyzing the hundreds of millions of electronic patient records already in existence; there is far more information to be gleaned there than in medical journals.
I think it rearranged the wavefront a bit, which could be seen as a "partial collapse". There may have also been some times when certain photons were localized to one slit, but that would have been experimental noise rather than what they were trying to detect.
Short answer: who says ability is less heritable in Sweden? And even if so, who says that Sweden allows higher ability people to keep their earnings, or even tolerates the idea that some people just have more general ability than others?
Long answer:
I'm not sure that the correlation is actually much less, rather the range of incomes is more restricted due to taxation and culture (not rewarding greater ability, denying its existence, or even ostracizing high ability people for insufficient humility). Also the Swedes are a more homogenous group than Americans, so the outcomes would also be expected to be more homogenous. OTOH, counter-intuitively, I believe the slighter differences might in some cultures actually lead to higher heritability of incomes (so long as we normalize to within-group z-scores), since it would take less of a shift in ability to change the outcome. But if the distribution of incomes is artificially narrowed relative to the distribution of abilities, as in Sweden, the reverse should happen. In America, with a winner-take-all mindset, of course you will see a wider spread of incomes, even after taking into account the more diverse population. And, as I said before, heritability isn't simply genetics - environments are also heritable, and they are also more diverse in America.
Yup, a good algorithm and strong key length are indispensable, but that's maybe 5% at most of what is needed for security. The algorithm per se is rarely the weak link - it's the implementation (there are always more wrinkles than anybody can foresee when writing such code), or the client computer with no physical security, a gazillion processes with more access than is safe, broken antivirus and firewall, insecure OS, poor configuration, browser vulnerabilities etc. ad nauseam.
Dropping SHA-bignum into an existing web application is like putting a bank-vault door into a camping tent.