| And, really, it's right to be quite skeptical of any scientist who argues that "my theory wasn't wrong, 15 years of data was all wrong". That's an extraordinary claim.
Indeed, but that's not the claim.
And it's quite right be be skeptical of any internet poster who is "skeptical" of results which are disliked who doesn't really understand where they came from but is smug about their skepticism.
Closer to actual facts:
a) there are no calibrated ground-based stations in the Arctic because there is no ground in the Arctic. b) the physics of increased greenhouse effect predicts larger effects in polar regions. (This also distinguishes global warming from enhanced greenhouse from global warming from increased solar output). c) global warming is, of course global d) combine (b) and (c) and you recognize that accurate quantification of global warming requires good evaluation of polar temperatures. e) previous temperature reconstructions used simple extrapolation or ignored Arctic regions with no data. f) authors propose new technique to assimilate data from multiple sources like satellites to improve coverage g) authors calibrate/validate technique where good data were known h) authors run the method and apply to Arctic regions with authentic missing data i) results show substantial warming larger than estimates previously used in (e). j) results with substantial warming in Arctic are more consistent with estimates using first-principles physics of greenhouse effects and what mainstream climate scientists have been predicting since 1992 or so.
next up:
k) scientists doing improved data processing showing closer correspondence to physics get accused of being shrill anti-capitalist nazis or the like.
Remember the previously skeptical Berkeley statistics professor---a favorite of the usual "skeptical" right-wing deception machine---who was convinced that the climatologists were doing their data analysis wrong and showing excessive global warming. And he & students got the underlying data sets and worked for years. And they found that the climatologists had the right answer all along (in fact their own estimates of warming were a touch higher than the climatologists).
Of course Mathematica's handling of functions and lists is good, it's nearly Lisp over again.
| There was a simple example where I tried to calculate a Bayes factor, and the expression was something like (1 - x)/(1 - y), where x and y were very small positive numbers, somewhere around the order of 10^-15. This calculation totally failed in R--the answer given was 1.
Is this a failure, or that 1 is the best approximation. Did you want (1-x)/(1-y) - 1 instead ? Then
Mathematica is not good for numerical computations---in practice people often use Matlab or R or Python as prototypes for software which eventually ends up being programmed in C++ or Fortran (or Java, occasionally). Getting good answers with finite-precision floating point is valuable.
| Besides rampant attention whores without real friends, people incapable of extracting their noses from the person in fronts arserhole, people who believe anyone who doesn't have a fffacebook page must worship at the evil Google Altar, and, "intelligence" agencies.
Sounds like there will be customers for decades.
Nobody ever lost a dollar by underestimating the taste of the American public. -- P.T. Barnum
It has tons of plans. Run by US Office of Personnel Management. For my state it had far more options than ACA. And you can look for dental and vision too.
During the ACA debate it was proposed that other people get access to these plans & system. Republicans vehemently opposed it.
"Will people will realize this web site is a screaming endorsement of Libertarianism: A few little guys beat out the whole US Government, again."
You mean searching the plans which will now take you despite pre-existing medical conditions?
Previously, plans were all sufficiently different that they worked as confusopoly to preclude clear competitive substitution, and they wouldn't take you, and the post-hoc dropped people who had developed serious medical problems because of trivial mistakes on the application years prior---and this was intentional policy to search for those when facing claims.
I believe that some state legislatures have prohibited employees from cooperating or implementing exchanges. Presumably any IT worker in a state datacenter would be fired, and possibly sued or prosecuted, if they worked to connect to healthcare.gov.
Kentucky strangely enough is going for it, except they make sure to mention the Federal government nowhere, so that its voters honestly think it was a Kentucky program..
Actually it's possible. The Federal Reserve is the checking account for the US Government (I mean in a literal sense, the Fed provides cash management & accounts for US Departments). The 'black budget' is not disclosed and it's not clear anybody in the Fed has the clearance to know about it. A man in sunglasses could order some clerk at the Fed to insert a certain number in a certain account and that's it. The Fed doesn't report the deficit, the Treasury department does and it works for the President.
At a minimum, CIA regularly gets substantial amount of physical US currency for its overseas operations. What's the bank for that?
FDIC insures bank's depositors against malefesance by the bank.
Congress & the Treasury Secretary insures banks, but only if you're part of the in crowd like Goldman Sachs, and JP Morgan, and not Bear Stearns/Lehman Bros or thousands of less cool banks. And you don't even have to pay large premiums upfront---dinners, jobs and lobbying is so much cheaper than premiums.
The commissioner of the FDIC during the crisis was pretty strenuously against insuring banks and look what it got her.
Much less likely to be oil industry, and much more likely to be financial institutions shorting the stock.
The threat to oil industry is slow and decades away---to them the problem is access to high quality oil fields currently held by nations and capital costs for fracking.
By contrast a 2 week hype/whinge cycle is perfect for a hedge fund.
|Who'd buy a company that's lost money over its entire existence and is only making operating profit on old products that it's about to replace?
Somebody who doesn't hire Microsoft middle managers to run things and finds the existing assets sufficiently valuable at the purchase price.
BTW: Elop is a terrible choice: Ballmer's ego and blindness, without Ballmer's brains.
Ray Ozzie---the one visionary Ballmer forced out----is a good alternative. Anybody who was successful in rising through Microsoft's corporate culture (e.g. Elop) is not the answer---because destroying that and remaking it again differently is the key.
SpaceX has the goal to be a successful orbital rocketry business.
These guys have the goal to noodle around and blow stuff up. Their capability is somewhere in the 1940's. Note "suborbital" << "orbital". You can go up to be 'suborbital'. If you want orbital you have to go up but more importantly around very very fast. Orbital velocity is about 7 km/s. The plot on these guys's page shows velocity topping out at 0.7 km/s. So they have 1/100th the energy needed for orbit (and obviously they have no capability to do multi-staging which is quite non-trivial to do reliably). And if they put more fuel in, then the mass goes up even more. It's just useless hobby waste.
Thing about rocket science is that SpaceX knows it's rocket science and employs people who know the scientific and engineering experience of 50 years of rocketry. They know they actually need substantial simulation and material science experimentation. They know they need to build a rocket engine test stand and understand fundamental dynamics in many regimes including near vacuum.
Modeling of fluid flow instabilities is far more quantitative today with large-scale simulations.
There are something called "dielectric barrier junction" plasma emitters which are starting to be used to manipulate the airflow at the boundary layer, and perhaps with some tricks they can reduce the instabilities.
The logical scenario is that they are developing it in secret and it has a number of test programs behind it...and now they want a bunch more money to make a number for real---and buried in the funding will be money for other R&D for classified drones.
Back in U-2/SR-71 days, they had enough funding at leisure to not need to make PR to get money.
It's not mathematical models----it's specific physical models which are represented in mathematics, combined with the empirically powerful knowledge of universal physical law.
FItting an arbitrary model with many degrees of freedom to one data set? yeah extrapolation is worrisome.
Extrapolating consequences of fundamental microscopic laws of physics interlinked and verified by millions of experiments and observations over a century of human civilization? No other extrapolation in the known human world is as useful or as secure.
Some states succeeded with their websites. The federal government succeeded with its employee insurance marketplace which has much wider coverage. http://www.opm.gov/healthcare-insurance/healthcare/
Republicans refused to allow people onto this plan, or to buy into Medicare.
ACA is not designed to fail intentionally but it probably will because it only addresses one part of a profit-making system. There is no competitive substitutability or clarity on prices (not just costs!). Ever try to find out how much some thing will cost at office X vs Y, with insurance? It's astonishingly difficult. I suspect this is intentional.
Single-payer appears to be empirically more successful for medicine (and few other goods and services).
Slashdot Mirror
| And, really, it's right to be quite skeptical of any scientist who argues that "my theory wasn't wrong, 15 years of data was all wrong". That's an extraordinary claim.
Indeed, but that's not the claim.
And it's quite right be be skeptical of any internet poster who is "skeptical" of results which are disliked who doesn't really understand where they came from but is smug about their skepticism.
Closer to actual facts:
a) there are no calibrated ground-based stations in the Arctic because there is no ground in the Arctic.
b) the physics of increased greenhouse effect predicts larger effects in polar regions. (This also distinguishes global warming from enhanced greenhouse from global warming from increased solar output).
c) global warming is, of course global
d) combine (b) and (c) and you recognize that accurate quantification of global warming requires good evaluation of polar temperatures.
e) previous temperature reconstructions used simple extrapolation or ignored Arctic regions with no data.
f) authors propose new technique to assimilate data from multiple sources like satellites to improve coverage
g) authors calibrate/validate technique where good data were known
h) authors run the method and apply to Arctic regions with authentic missing data
i) results show substantial warming larger than estimates previously used in (e).
j) results with substantial warming in Arctic are more consistent with estimates using first-principles physics of greenhouse effects and what mainstream climate scientists have been predicting since 1992 or so.
next up:
k) scientists doing improved data processing showing closer correspondence to physics get accused of being shrill anti-capitalist nazis or the like.
Remember the previously skeptical Berkeley statistics professor---a favorite of the usual "skeptical" right-wing deception machine---who was convinced that the climatologists were doing their data analysis wrong and showing excessive global warming. And he & students got the underlying data sets and worked for years. And they found that the climatologists had the right answer all along (in fact their own estimates of warming were a touch higher than the climatologists).
"social-science abomination like R with a screwball syntax."
R is a clone of the S+ language, which was invented at Bell Labs. SPSS is the social-science 'abomination'.
Of course Mathematica's handling of functions and lists is good, it's nearly Lisp over again.
| There was a simple example where I tried to calculate a Bayes factor, and the expression was something like (1 - x)/(1 - y), where x and y were very small positive numbers, somewhere around the order of 10^-15. This calculation totally failed in R--the answer given was 1.
Is this a failure, or that 1 is the best approximation. Did you want (1-x)/(1-y) - 1 instead ? Then
Mathematica is not good for numerical computations---in practice people often use Matlab or R or Python as prototypes for software which eventually ends up being programmed in C++ or Fortran (or Java, occasionally). Getting good answers with finite-precision floating point is valuable.
El Al has about fifty international flights per day (not much point to domestic air inside Israel). The USA has tens of thousands.
El Al can hire high-end, experienced intelligence operatives for this task. TSA obviously doesn't.
http://usatoday30.usatoday.com/news/attack/2001/10/01/elal-usat.htm#more
They can spend 10 minutes asking question to half the passengers of each flight.
What do you call a user who doesn't know how Jeopardy works: the user provides the answers and I ask the questions?
| Besides rampant attention whores without real friends, people incapable of extracting their noses from the person in fronts arserhole, people who believe anyone who doesn't have a fffacebook page must worship at the evil Google Altar, and, "intelligence" agencies.
Sounds like there will be customers for decades.
Nobody ever lost a dollar by underestimating the taste of the American public. -- P.T. Barnum
This one works.
http://www.opm.gov/healthcare-insurance/healthcare/
It has tons of plans. Run by US Office of Personnel Management. For my state it had far more options than ACA. And you can look for dental and vision too.
During the ACA debate it was proposed that other people get access to these plans & system. Republicans vehemently opposed it.
"Will people will realize this web site is a screaming endorsement of Libertarianism: A few little guys beat out the whole US Government, again."
You mean searching the plans which will now take you despite pre-existing medical conditions?
Previously, plans were all sufficiently different that they worked as confusopoly to preclude clear competitive substitution, and they wouldn't take you, and the post-hoc dropped people who had developed serious medical problems because of trivial mistakes on the application years prior---and this was intentional policy to search for those when facing claims.
That was the libertarian paradise.
I believe that some state legislatures have prohibited employees from cooperating or implementing exchanges. Presumably any IT worker in a state datacenter would be fired, and possibly sued or prosecuted, if they worked to connect to healthcare.gov.
Kentucky strangely enough is going for it, except they make sure to mention the Federal government nowhere, so that its voters honestly think it was a Kentucky program..
Actually it's possible. The Federal Reserve is the checking account for the US Government (I mean in a literal sense, the Fed provides cash management & accounts for US Departments). The 'black budget' is not disclosed and it's not clear anybody in the Fed has the clearance to know about it. A man in sunglasses could order some clerk at the Fed to insert a certain number in a certain account and that's it. The Fed doesn't report the deficit, the Treasury department does and it works for the President.
At a minimum, CIA regularly gets substantial amount of physical US currency for its overseas operations. What's the bank for that?
| So what I don't get is why people would keep their bitcoins in one of these "banks" at all.
So they can trade bitcoins frequently.
FDIC insures bank's depositors against malefesance by the bank.
Congress & the Treasury Secretary insures banks, but only if you're part of the in crowd like Goldman Sachs, and JP Morgan, and not Bear Stearns/Lehman Bros or thousands of less cool banks. And you don't even have to pay large premiums upfront---dinners, jobs and lobbying is so much cheaper than premiums.
The commissioner of the FDIC during the crisis was pretty strenuously against insuring banks and look what it got her.
|Do these bookstores really think that refusing to sell the devices themselves will slow adoption?
Of their regular customers? You betcha.
Just reminding them about their competitors' product isn't a good idea.
Much less likely to be oil industry, and much more likely to be financial institutions shorting the stock.
The threat to oil industry is slow and decades away---to them the problem is access to high quality oil fields currently held by nations and capital costs for fracking.
By contrast a 2 week hype/whinge cycle is perfect for a hedge fund.
Looking for a few good geeks, and a whole bunch o' bad ones.
|Who'd buy a company that's lost money over its entire existence and is only making operating profit on old products that it's about to replace?
Somebody who doesn't hire Microsoft middle managers to run things and finds the existing assets sufficiently valuable at the purchase price.
BTW: Elop is a terrible choice: Ballmer's ego and blindness, without Ballmer's brains.
Ray Ozzie---the one visionary Ballmer forced out----is a good alternative. Anybody who was successful in rising through Microsoft's corporate culture (e.g. Elop) is not the answer---because destroying that and remaking it again differently is the key.
SpaceX has the goal to be a successful orbital rocketry business.
These guys have the goal to noodle around and blow stuff up. Their capability is somewhere in the 1940's. Note "suborbital" << "orbital". You can go up to be 'suborbital'. If you want orbital you have to go up but more importantly around very very fast.
Orbital velocity is about 7 km/s. The plot on these guys's page shows velocity topping out at 0.7 km/s. So they have 1/100th the energy needed for orbit (and obviously they have no capability to do multi-staging which is quite non-trivial to do reliably). And if they put more fuel in, then the mass goes up even more. It's just useless hobby waste.
Thing about rocket science is that SpaceX knows it's rocket science and employs people who know the scientific and engineering experience of 50 years of rocketry. They know they actually need substantial simulation and material science experimentation. They know they need to build a rocket engine test stand and understand fundamental dynamics in many regimes including near vacuum.
It probably makes much too much nitrogen oxide smog molecules to be legal.
Modeling of fluid flow instabilities is far more quantitative today with large-scale simulations.
There are something called "dielectric barrier junction" plasma emitters which are starting to be used to manipulate the airflow at the boundary layer, and perhaps with some tricks they can reduce the instabilities.
The logical scenario is that they are developing it in secret and it has a number of test programs behind it...and now they want a bunch more money to make a number for real---and buried in the funding will be money for other R&D for classified drones.
Back in U-2/SR-71 days, they had enough funding at leisure to not need to make PR to get money.
It's not mathematical models----it's specific physical models which are represented in mathematics, combined with the empirically powerful knowledge of universal physical law.
FItting an arbitrary model with many degrees of freedom to one data set? yeah extrapolation is worrisome.
Extrapolating consequences of fundamental microscopic laws of physics interlinked and verified by millions of experiments and observations over a century of human civilization? No other extrapolation in the known human world is as useful or as secure.
Some states succeeded with their websites. The federal government succeeded with its employee insurance marketplace which has much wider coverage.
http://www.opm.gov/healthcare-insurance/healthcare/
Republicans refused to allow people onto this plan, or to buy into Medicare.
ACA is not designed to fail intentionally but it probably will because it only addresses one part of a profit-making system. There is no competitive substitutability or clarity on prices (not just costs!). Ever try to find out how much some thing will cost at office X vs Y, with insurance? It's astonishingly difficult. I suspect this is intentional.
Single-payer appears to be empirically more successful for medicine (and few other goods and services).
| I found myself thinking, "So? What's new about that?"
Quantification, a reasonably precise definition of predictive power, and empirical/observational results.
"That is why science wins eventually and every time over superstition and ignorance."
Unless accompanied by massive barbarian hordes.
"Non-science can only win if no players remain
That's an accepted strategy: off with their head.