Some interesting perspective from Matt Strassler, who's a particle physicist at Harvard.
He points out that this is still an *indirect* observation of gravitational waves (and not the first one) and that the results look sensibly in line with some predictions from inflation. And that while this is a tremendous experiment, it's not any kind of "smoking gun", and we really need to wait for replication to get properly excited.
I'm still trying to find a way to use the GPU for computations without having to jump through crazy hoops to do it. Also, multithreading in general is often a PITA to get right...
This.
It still seems like it's really hard to write maintainable, extendable code that does parallel execution in anything more than a trivial way (break up a calc into independent pieces with no interaction).
Does anyone know if there's been any interesting work on new language paradigms for this sort of thing? It feels like the community is waiting for some kind of change in semantics that'd make this easier for the bulk of developers who aren't parallel execution gurus. Like object oriented coding was for the ability to scale code by developing intelligible interfaces: not something you couldn't do with earlier paradigms, but it made it easier.
"NASA's FY 2008 budget of $17.3 billion represents about 0.6% of the $2.9 trillion United States federal budget."
So a bit more than 0.02%:) but still pretty small. I guess you're claiming something about the NASA budget ex-ISS, which I don't know, but I suspect that less than 96% of the total NASA budget is spent on the ISS.
Since the Pharma industry is based heavily on biology and bio-chemistry and in turn on theories of evolution, maybe we could start a campaign to equate medical drug use with support of evolution. Hit the zealots where they live (literally) by accusing them of supporting, by act, the theory of evolution if they take any medical drugs. Suggest if they really do not support evolution, they should forgo their medicine.
As much as I don't like the intelligent design folks, this is not a valid argument. ID doesn't say that evolution *never* happens - that would be a truly ridiculous claim, since there's direct evidence of it all over the place in the bacterial and viral worlds. Instead, they say that evolution can't explain *all* diversity, and that the big steps were magically instigated by a higher power.
So an ID person can quite happily live with drug design based on evolutionary principles, since they believe low-level evolution does happen.
Nothing hurts the argument against ID more than making an invalid point - they jump on it and demolish it, and make their side look stronger in the process to people who don't know what to believe. Remember that most of these people don't care what the truth is, they just want to win the argument, and they'll play dirty to do it.
There are much more dangerous illnesses than SARS which affect much more people.
The issue with SARS isn't how many people are currently sick - it's how many *could* be sick. In the 20s a flu epidemic swept through Europe, killing 2MM people. Sounds like a lot... but it only had a 2.5% fatality rate. The problem was that everyone caught it.
Same deal with SARS. Relatively low fatality rate (4-6%, maybe as high as 10%), so it doesn't kill people before they spread it (which is easy to do, it seems), but high enough that it can do some damage if a lot of people catch it.
Hopefully the world community has reacted fast enough with quarantines, etc, to stamp it out... but if it starts spreading freely in the wild, a lot of people's lives are in danger.
Not at Waterloo - it's part of the degree program. A normal 4-year undergrad runs 5 years, with a bunch of "work terms" thrown into the mix. They're organized through the university.
As provocative as the book's thesis seems to be -- and I must admit that my information on it comes entirely from this review -- it's not new. In the 1970s, Burton Malkiel's A Random Walk Down Wall Street posited that market fluctuations are mutually independent, though the market follows a general upwards trend, and thus it's impossible (ceteris paribus) to make any bets on short-term stock performance.
That's a bit inaccurate in two ways:
1) Taleb doesn't claim the market is an efficient random walk - he thinks the market inefficiently prices events that have low probability because people are bad at estimating low-prob events. He capitalizes on that mispricing by buying low-prob events, mostly losing money but occasionally scoring big (and hopefully making money on average). 2) Even if the market *is* a random walk, you can still expect to make money in the long term because the stock market pays a better return on average than, say, the bond market. That's because investors demand a higher return in exchange for the extra risk due to stock volatility. But that only applies to investing for the long term, not day trading. Over short terms the extra return is completely swamped by the noise from volatility.
After reading the announcement and a couple of the links off it, this sounds more like an XML standard that law enforcement agencies and legal departments can use to send each other information during an investigation - not something they'd foist on the general public to make it easier for government to spy on us.
My first impressions after twenty minutes of playing against bots: the "feel" is a lot more like Q3 than the original UT. That's a little disappointing for me.
Everything considered, the chances of being killed by an asteroid are considered to be about one in 12,000, compared to the one in 10,000 possibility of being killed in an airline crash, he said.
Errr... wtf? The odds of my dying in an airline crash are much less than 1 in 10,000, thank you very much. Ditto for being whacked by an asteroid.
Seems like they're off by a bunch of orders of magnitude.
According to this the odds are about one in four *million*.
Like Sagan says, people laughed at the Wright brothers; but they also laughed at the Marx brothers. Extraordinary claims require extraodinary proof, and nothing in those links does it for me.
what's the difference between a scientist who is a Christian and one who believes in other unprovable, irrational propositions such as clairvoyance or astrology?
The difference: religious beliefs are beliefs about metaphysics (does god exist? etc) which are outside the domain of science because they are fundamentally unproveable.
Psychic ability, on the other hand, claims to be empirically testable! Unfortunately, whenever any of these "abilities" are put to the test they don't pass.
You can't put religion to a test, but you can certainly put psychic powers and UFOs to a test. That's the difference.
strange matter
on
Quark Stars
·
· Score: 4, Interesting
I did a PhD on pulars, which everyone thinks are neutron stars. At one point I found a paper which suggested that instead they might be "strange matter" stars - and it's always intrigued me how difficult it is to distinguish between the two.
The cool thing about finding strange matter stars is that it suggests there's a lower-energy state of matter than our normal up/down quark pairings. No one's really sure because QCD is so hard to get numbers out of.
Every time they build a new accelerator someone harps on this, worrying about whether we'll ram particles together hard enough to create a meta-stable bubble of strange matter. If there is a net saving in energy due to expanding that bubble (drop in energy due to increasing volume of lower-energy-state matter, increase in energy due to increased surface tension on the surface), the bubble will tend to expand and gobble up everything in its path - like the Earth, for example.
That's the common worry, though it's easily allayed by noting that particles with much higher energy than anything we could create in an accelerator are hitting our atmosphere all the time, and none of them have turned our planet into a jiggling mass of strange matter.
Good thing Slashdot is here to give us unbiased journalism.
Arg - why do people keep trying to bring this up? Slashdot *isn't* journalism - it's a place where you and I can *discuss* journalism from other sources.
The real products Red Hat sells are convenience (get everything in one place with manuals) and support.
Those products really are worth money, even though the underlying software isn't (since it's Free). It's not too unbelievable that you can support a business on them. Convenience, though, is probably less valuable - especially as bandwidth gets cheaper and download times get shorter.
If all Sun is doing with StarOffice is selling convenience - and not support - then I suspect they'll have a hard time of it in the long run. If they're including support it doesn't sound as bad.
I did undergrad in engineering physics, PhD in astrophysics, then found out that pursuing a professorship is a difficult proposition.
So I switched gears entirely and took up quantitative finance, and I'm now working on a trading desk doing modeling and risk management.
Kind of an unusual route, but that's the advantage of physics - it gives you a broad background in math and problem-solving that you can apply to lots of different fields.
Other people in my engineering physics class have gone on to do aerospace engineering, law, business school, programming, architecture, and lots of other stuff.
So: do physics. It's fun, suitably geeky, and it opens lots of doors for the future.
Yeah because prohibiting the export of this will prevent anyone evil from getting hold of it...
As much as I'm against encryption controls, this argument is easily refuted by noting that, in this case, the export controls *did* cause this particular bad guy to use weaker encryption.
Maybe sophisticated terrorists could get around export controls trivially - but most of them probably aren't terribly sophisticated.
Number 1 is the meteor problem, during the 2 wks in which the moon leads the earth through space, the combined gravational forces of the moon and earth significantly increase the probability of a meteor striking the telescope system if it where built on the far side
How did this get modded up to 5, Insightful?? This is totally ridiculous. a) the moon has a tiny gravitational field compared to the Earth, b) the area of the moon is tiny compared to that of the Earth - it's not going to stop an appreciable amount of meteors. The reason the Earth isn't cratered is because there aren't that many meteors anymore (compared to 3B years ago), and because water/plant life smoothes out impact craters in a relatively short period of time.
I think someone's watched Armageddon a few too many times.
The second major problem is that over half the time the telescope would be pointed at or at least exposed to the sun which in it self is a significant source of rfi.
Kind of like radio telescopes on the Earth, you mean? How could anyone do any radio astronomy on the Earth with that annoying Sun there??
> By extention it doesn't follow that there is neccesarily a reason for the big bang being true or not true. There doesn't strictly have to be any explanation for why it is the way it is.
That's not what Godel's incompleteness theorem says at all. It just says that in any sufficiently complex system there are statements which you can't *prove* are true or false.
They're still either true or false; it'd just take an infinite amount of time with a computer to calculate the truth value, so we can't do the calculation.
Slashdot Mirror
Some interesting perspective from Matt Strassler, who's a particle physicist at Harvard.
He points out that this is still an *indirect* observation of gravitational waves (and not the first one) and that the results look sensibly in line with some predictions from inflation. And that while this is a tremendous experiment, it's not any kind of "smoking gun", and we really need to wait for replication to get properly excited.
> Can anyone find a free copy that we can examine?
Archive link
I'm still trying to find a way to use the GPU for computations without having to jump through crazy hoops to do it. Also, multithreading in general is often a PITA to get right...
This.
It still seems like it's really hard to write maintainable, extendable code that does parallel execution in anything more than a trivial way (break up a calc into independent pieces with no interaction).
Does anyone know if there's been any interesting work on new language paradigms for this sort of thing? It feels like the community is waiting for some kind of change in semantics that'd make this easier for the bulk of developers who aren't parallel execution gurus. Like object oriented coding was for the ability to scale code by developing intelligible interfaces: not something you couldn't do with earlier paradigms, but it made it easier.
> makes an analogy with the (marginalized, demonized) economists who warned of a coming economic collapse in 2007.
Hrm... like Paul Krugman, the economic who warned of collapse and yet received a Nobel prize?
From the NASA budget wiki page:
:) but still pretty small. I guess you're claiming something about the NASA budget ex-ISS, which I don't know, but I suspect that less than 96% of the total NASA budget is spent on the ISS.
"NASA's FY 2008 budget of $17.3 billion represents about 0.6% of the $2.9 trillion United States federal budget."
So a bit more than 0.02%
Since the Pharma industry is based heavily on biology and bio-chemistry and in turn on theories of evolution, maybe we could start a campaign to equate medical drug use with support of evolution. Hit the zealots where they live (literally) by accusing them of supporting, by act, the theory of evolution if they take any medical drugs. Suggest if they really do not support evolution, they should forgo their medicine.
As much as I don't like the intelligent design folks, this is not a valid argument. ID doesn't say that evolution *never* happens - that would be a truly ridiculous claim, since there's direct evidence of it all over the place in the bacterial and viral worlds. Instead, they say that evolution can't explain *all* diversity, and that the big steps were magically instigated by a higher power.
So an ID person can quite happily live with drug design based on evolutionary principles, since they believe low-level evolution does happen.
Nothing hurts the argument against ID more than making an invalid point - they jump on it and demolish it, and make their side look stronger in the process to people who don't know what to believe. Remember that most of these people don't care what the truth is, they just want to win the argument, and they'll play dirty to do it.
is that considered food? if not, then i can survive without food for 5 day easily. :)
I think beef jerky safely falls into the old "eatable but not edible" category.
That's an interesting question: in the field of language comprehension, is the cutting edge of research in academia or in industry?
Anyone know?
"How on earth do you test wether or not this vaccine has worked?"
Animal trials. They already did them, and none of the exposed animals contracted Ebola.
There are much more dangerous illnesses than SARS which affect much more people.
The issue with SARS isn't how many people are currently sick - it's how many *could* be sick. In the 20s a flu epidemic swept through Europe, killing 2MM people. Sounds like a lot... but it only had a 2.5% fatality rate. The problem was that everyone caught it.
Same deal with SARS. Relatively low fatality rate (4-6%, maybe as high as 10%), so it doesn't kill people before they spread it (which is easy to do, it seems), but high enough that it can do some damage if a lot of people catch it.
Hopefully the world community has reacted fast enough with quarantines, etc, to stamp it out... but if it starts spreading freely in the wild, a lot of people's lives are in danger.
Not at Waterloo - it's part of the degree program. A normal 4-year undergrad runs 5 years, with a bunch of "work terms" thrown into the mix. They're organized through the university.
As provocative as the book's thesis seems to be -- and I must admit that my information on it comes entirely from this review -- it's not new. In the 1970s, Burton Malkiel's A Random Walk Down Wall Street posited that market fluctuations are mutually independent, though the market follows a general upwards trend, and thus it's impossible (ceteris paribus) to make any bets on short-term stock performance.
That's a bit inaccurate in two ways:
1) Taleb doesn't claim the market is an efficient random walk - he thinks the market inefficiently prices events that have low probability because people are bad at estimating low-prob events. He capitalizes on that mispricing by buying low-prob events, mostly losing money but occasionally scoring big (and hopefully making money on average).
2) Even if the market *is* a random walk, you can still expect to make money in the long term because the stock market pays a better return on average than, say, the bond market. That's because investors demand a higher return in exchange for the extra risk due to stock volatility. But that only applies to investing for the long term, not day trading. Over short terms the extra return is completely swamped by the noise from volatility.
After reading the announcement and a couple of the links off it, this sounds more like an XML standard that law enforcement agencies and legal departments can use to send each other information during an investigation - not something they'd foist on the general public to make it easier for government to spy on us.
I see where you're coming from, but I've played a fair bit of both UT and Q3, and though it's hard to put my finger on why, I much prefer UT.
My first impressions after twenty minutes of playing against bots: the "feel" is a lot more like Q3 than the original UT. That's a little disappointing for me.
Everything considered, the chances of being killed by an asteroid are considered to be about one in 12,000, compared to the one in 10,000 possibility of being killed in an airline crash, he said.
Errr... wtf? The odds of my dying in an airline crash are much less than 1 in 10,000, thank you very much. Ditto for being whacked by an asteroid.
Seems like they're off by a bunch of orders of magnitude.
According to this the odds are about one in four *million*.
This seems to be a pretty extraordinary claim!
Like Sagan says, people laughed at the Wright brothers; but they also laughed at the Marx brothers. Extraordinary claims require extraodinary proof, and nothing in those links does it for me.
what's the difference between a scientist who is a Christian and one who believes in other unprovable, irrational propositions such as clairvoyance or astrology?
The difference: religious beliefs are beliefs about metaphysics (does god exist? etc) which are outside the domain of science because they are fundamentally unproveable.
Psychic ability, on the other hand, claims to be empirically testable! Unfortunately, whenever any of these "abilities" are put to the test they don't pass.
You can't put religion to a test, but you can certainly put psychic powers and UFOs to a test. That's the difference.
I did a PhD on pulars, which everyone thinks are neutron stars. At one point I found a paper which suggested that instead they might be "strange matter" stars - and it's always intrigued me how difficult it is to distinguish between the two.
The cool thing about finding strange matter stars is that it suggests there's a lower-energy state of matter than our normal up/down quark pairings. No one's really sure because QCD is so hard to get numbers out of.
Every time they build a new accelerator someone harps on this, worrying about whether we'll ram particles together hard enough to create a meta-stable bubble of strange matter. If there is a net saving in energy due to expanding that bubble (drop in energy due to increasing volume of lower-energy-state matter, increase in energy due to increased surface tension on the surface), the bubble will tend to expand and gobble up everything in its path - like the Earth, for example.
That's the common worry, though it's easily allayed by noting that particles with much higher energy than anything we could create in an accelerator are hitting our atmosphere all the time, and none of them have turned our planet into a jiggling mass of strange matter.
Anyway, interesting idea.
Good thing Slashdot is here to give us unbiased journalism.
Arg - why do people keep trying to bring this up? Slashdot *isn't* journalism - it's a place where you and I can *discuss* journalism from other sources.
why does the same logic not apply?
The real products Red Hat sells are convenience (get everything in one place with manuals) and support.
Those products really are worth money, even though the underlying software isn't (since it's Free). It's not too unbelievable that you can support a business on them. Convenience, though, is probably less valuable - especially as bandwidth gets cheaper and download times get shorter.
If all Sun is doing with StarOffice is selling convenience - and not support - then I suspect they'll have a hard time of it in the long run. If they're including support it doesn't sound as bad.
I did undergrad in engineering physics, PhD in astrophysics, then found out that pursuing a professorship is a difficult proposition.
So I switched gears entirely and took up quantitative finance, and I'm now working on a trading desk doing modeling and risk management.
Kind of an unusual route, but that's the advantage of physics - it gives you a broad background in math and problem-solving that you can apply to lots of different fields.
Other people in my engineering physics class have gone on to do aerospace engineering, law, business school, programming, architecture, and lots of other stuff.
So: do physics. It's fun, suitably geeky, and it opens lots of doors for the future.
Yeah because prohibiting the export of this will prevent anyone evil from getting hold of it...
As much as I'm against encryption controls, this argument is easily refuted by noting that, in this case, the export controls *did* cause this particular bad guy to use weaker encryption.
Maybe sophisticated terrorists could get around export controls trivially - but most of them probably aren't terribly sophisticated.
Number 1 is the meteor problem, during the 2 wks in which the moon leads the earth through space, the combined gravational forces of the moon and earth significantly increase the probability of a meteor striking the telescope system if it where built on the far side
How did this get modded up to 5, Insightful?? This is totally ridiculous. a) the moon has a tiny gravitational field compared to the Earth, b) the area of the moon is tiny compared to that of the Earth - it's not going to stop an appreciable amount of meteors. The reason the Earth isn't cratered is because there aren't that many meteors anymore (compared to 3B years ago), and because water/plant life smoothes out impact craters in a relatively short period of time.
I think someone's watched Armageddon a few too many times.
The second major problem is that over half the time the telescope would be pointed at or at least exposed to the sun which in it self is a significant source of rfi.
Kind of like radio telescopes on the Earth, you mean? How could anyone do any radio astronomy on the Earth with that annoying Sun there??
I suspect we could live with this.
> By extention it doesn't follow that there is neccesarily a reason for the big bang being true or not true. There doesn't strictly have to be any explanation for why it is the way it is.
That's not what Godel's incompleteness theorem says at all. It just says that in any sufficiently complex system there are statements which you can't *prove* are true or false.
They're still either true or false; it'd just take an infinite amount of time with a computer to calculate the truth value, so we can't do the calculation.