How many times have you asked someone, "What does your code do to solve this problem?" and got a description of an algorithm which, when you finally get to see the source, does not match the code?
In my case, the answer to that question is, "Lots." I have had it happen in pure science (neutrino physcis), applied science (medical physics) and software development (database programming, data analysis, etc.)
I am painfully aware that my own published descriptions of algorithms have often left out minor details that may be critical in some applications, but that page limits in peer-reviewed journals necessitate. It is not uncommon to get a call from someone doing similar work asking for details about what you've done, how you've done it, and in some cases, asking to look at source code.
In contentious areas of science such requests are not always met with full disclosure, which is a sign that the people involved are no longer doing science. They are doing politics. This happens a lot, and it brings the scientific process to a halt on the question at issue.
In the case at hand, the original authors have done a very poor job of describing what they have done, and an extremely poor job of defending their work. Their refusal to publish their source code for their analysis gives credibility to their critics.
There are certainly legitimate cases where code ought not be published. If a lab has spent many, many years developing a framework for solving a certain type of problem and wants to get the most advantage out of that framework before releasing it, they may reasonably want to limit it's disemination for a while. But those sorts of reason don't apply in this case, and the source should be made available to anyone who wants to reproduce their actual results. That would just be good science.
The sphericity of the Earth was not generally accepted until much later than the 1st century C.E. More like the 8th. Even then, the way in which people thought about the Earth was radically different than the modern idea of a spherical Earth:
http://abob.libs.uga.edu/bobk/whiteb03.html
The "doctrine of the antipodes" asserted that even if the Earth was spherical, no humans lived on "the other side" because they'd have their feet in the air, wouldn't be able to observe the descent of Christ at the 2nd coming, etc.
Variants of this doctrine persisted long after the nominal sphericity debate had been settled, and I'd argue that until something like the modern view of a spherical Earth, antipodes and all, was generally accepted, it is not quite correct to claim that it was "generally accepted that the earth is round".
Ancient ideas are alien to our own, and it is easy to impose our modern understanding on the words the ancients used, creating great distortion. So I get to disagree with everyone: in the first millenium C.E. people neither believed that the Earth was flat, nor that the Earth was round in the modern sense. They believed the Earth had a special place in the universe, and their understanding of the shape and geography of the Earth grew out of Church doctrine as much as emprical observation.
The Dark Matter problem is actually the Dark Matter problemS, plural. Galactic dark matter is only the tip of the iceberg, and can be explained by baryonic matter. Dark matter (or energy) on larger scales is a different kettle of fish. A better heading would have been: "Large scale dynamics of the universe", which would take in the horizon problem, the dark matter problem and the dark energy problem.
The article in this regard is a bit like a software requirements document written by a user: it's in terms of projected solutions rather than actual problems. The actual problem is that we don't understand the large-scale dynamics of the universe. The solution may be anything from exotic particles to weird properties of space to alternative dynamics. We just don't know.
The things about alpha changing and tetra-neutrons are cool, but far more likely to be mistakes than new phenomena.
The stuff about high-energy cosmic rays is probably the most interesting, and in fact there are a wealth of phenomena that have been observed by large detectors such as SNO and Kamiokande over the years that really don't make sense. The possibility of new physics at high energies, or entirely novel particles (magnetic monopoles, for example) is quite real, and some of the anomalies observed in these detectors may be indicators of these things.
Some of us still do live in caves. And guess what: there are substantial gender inequalities in modern hunter-gatherer societies. Industrial societies have lower gender inequality than any society that has ever existed, in part because machines take the place of muscle, making the only significant difference between men and women irrelevant.
Women got the vote in the most heavily industrialized parts of the world first, and women's rights in the most heavily industrialized parts of the world are still far in advance of anywhere else.
Women in the industrial world can own property, enter into contracts, marry whom they please. This is not the case in many pre-industrial/developing societies in, for example, sub-Saharan Africa.
A surgeon of my acquaintance used to claim that there were only two significant medical advances in the past 200 years: anti-biotics, and sterile technique. To that I would add anesthetic, although that's more of an enabling technology than a cure. The effect of everything else has been tiny in comparison.
Probably the largest "minor" advance was the use of radiation to treat cancer. However, the success rate (60% of patients improve, 40% don't) has not changed significantly since the 50's, when Canadian doctors pioneered cobalt therapy. We've spent oodles of money on improvements since then, and some of those improvements have had a big effect for a small number of people, but the big numbers have not changed significantly.
The reality is that the largest increase in human health world-wide would be achieved by taking 100% of medical research dollars in the developed world and spending them on basic infrastructure--like clean water and sewage treatment--in the developing world.
What is the "fundamental rule of honest dealing" you are refering to? I am not aware of any rule that says I should not read a web page written to me, about me, that is trivially available to me.
The issue here is the triviality of availablity: the letters were available using the backspace and enter keys. Harvard made no attempt to restrict access other than by not linking them, which is not security. Many people, myself included, believe that if information about myself is freely accessible then there is nothing wrong with accessing it. An unlinked, unsecured page falls well within my definition of freely accessible--it differs from a paper letter left on someone's desk because it was not left on the website but rather put there deliberately, and should not have been put there if it was not meant to be seen.
The people who are morally culpable here are Harvard, for making information freely accessible to applicants and then punishing them for accessing it.
To forestall more invalid analogies: this is totally unlike walking into someone else's unlocked house, where there is a clear presumption in law and culture that I have no right to do so. This is not about picking up money on the sidewalk that I've seen fall out of someone's pocket and have clear knowledge of ownership. This is not like cheating on my wife: I am not breaking any promises I made (assuming Harvard applicants don't sign any undertaking that they will not attempt to access their admissions status by any means whatsoever.)
This is about information. Information about me. Information that is available to me by hitting the backspace and enter key.
Here's an analogy that might be valid: an applicant calls the admissions office and asks the person who answers the phone, "Am I in?" and the person who answers the phone, who happens to have the applicant's file in front of them, glances at it and says, "Looks like you're rejected."
Then that person tells a friend, who tells a friend, and pretty soon Harvard has 119 people calling them. Most of them get someone who says, "We don't know yet." Some get the dunderhead the first person talked to, who tells them "Looks like you're rejected." No one gets told "You're in."
Now, are these people who are calling the admissions office doing anything unethical? They are doing more work than the people who checked the URL online. How does the fact that Harvard had a poorly trained employee who made information available that should not have been an ethical problem for the people who tried to access that information?
Many of those who accessed it were probably doing the moral equivalent of calling the admissions office: they knew the official date hadn't arrived yet, but they thought they'd just check out this possible alternative and see. After all, if they shouldn't get the information, then Harvard would surely not give it to them.
"When you look at the non-industry sponsored research, it's about three to one--three out of every four papers shows an effect," Lai says. "Then, if you look at the industry-funded research, it's almost opposite--only one out of every four papers shows an effect." Ever try to get a null result published?
I believe that industry-sponsored research is biased. But simply because research is not industry-sponsored does not mean it is not biased.
In particular, what a scientist wants to see in an experiment is a positive effect, an non-null result. I've seen people (in genetics, as it happens) do terrible things to their data to get a non-null result, and carefully massage the statistics to make a result that deviates from the null hypothesis by a miniscule amount look significant.
Why?
Because it's a hell of a lot easier (to say nothing of more personally satisifying) to do all this work, kill all these rats, and at the end of the day be able to say something more interesting than, "Nothing to see here, move along..."
...3% false positives is still orders of magnitude too high for a deployable system. There are a range of interesting things they might do to improve the accuracy.
The natural move from my point of view is to look at mu-N interactions, where a muon blows apart a nucleus in the target material, producing a shower of excited nuclear fragments and neutrons. Heavy materials such as plutonium will have a much different cascade signature than relatively light things like iron, so it may be possible to develop a quite specific finger-printing mechanism that would be hard to work around. With a muon detector on top to act as a trigger, and some combination of gamma and neutron detectors nearby, this is might be able to both speed up processing and improve accuracy dramatically.
Of course, terrorists could always fall back to the obvious plan B: smuggling the weapon in hidden in a bale of marijuana.
They found that just about everyone could, on a small but repeatable level, affect the output of a random number generator just by concentrating on it. (The implications of that, if true, are staggering enough alone)
Extremely careful analysis is required when looking for very small effects in the midst of large masses of data.
Frequentist analysis breaks down in a variety of circumstances, and Bayesian analysis must be used instead. The most familiar case where frequentist analysis breaks down is when there are a very small number (or just one) event(s). But it also breaks down in these large datasets when one goes hunting for very small probability events.
Looked at informally, what is more probable: that humans have a small but significant ability to alter events by thinking about them (that evolution has somehow missed out on improving on) or that the experiments and analysis are somehow flawed? Naively, the latter hypothesis is more plausible, and the paper linked above demonstrates this to be the case.
In that case you have a specification! In the form of your static prototype.
A prototype and a specification do not contain the same information. A prototype consists of a single concrete instance of the thing a specification describes. It contains more information than the specification in some respects (the concrete design choices the implementor has made to fill in the gaps the specification is silent on) but more importantly it is also missing information that is absolutely required in a specification.
For example: What are the tolerances on particular processes? What are the scaling and responsiveness requirements? And so on.
Nothing that does not capture this kind of information is a specification, and so far as I know only a document can capture this kind of information.
Don't get me wrong: prototypes are useful. But they are not specifications, because they do not contain the all of the information that a specification must contain.
The key is he has top notch technical & marketing leaders on his team that guide the overall technical direction.
No, the key is that he listens to them and trusts them when they advise him on things he does not understand. Lots of companies have good technical advisors. Very few pay any attention to them.
De jure Canada has one-tier systems. De facto we are multi-tier, and becoming moreso. In education, kids who go to York or Carleton don't get the same quality of education as kids who go to Queen's or even U of T. And for the rich there are much better health care options available, including private clinics in the U.S.
The political mythology of a single-tier system has helped to maintain the quality of that system until the past few years, but the reality is that Canadians have more choices available than they are often willing to admit to. Nor are our taxes that much higher, and they are likely to get lower due to our more fiscally responsible governments.
Canadian high schools are better than their U.S. counter-parts, but it's my impression that local alternatives are a larger force here, including mixed schools where a number of different educational philosphies or foci are represented by different organizations within a single physical building. This diversity, both within and without the public system, is one of the things that is keeping our public schools from completely failing in their charge.
I saw a talk by a patent agent/IP lawyer that was basically FUD about FOSS, and one of his big points was, "There are more than forty different open source licenses out there!" He didn't mention that there are thousands of different closed source licenses, and somehow the business community has not ground to a halt while trying to deal with their complexity and diversity.
I think that might have been because people like him make their money out of all that complexity, and so the simplicity and directness of most open source licenses are a profound threat to their income. We are about as likely to see a standard set of closed source licenses as we are to see a national standard building code, and for the same reason: there are powerful economic interests opposed to them.
30 kg of plutonium is more than enough. There are reprocessing issues even with plutonium, though--you need to get rid of the most of the relatively short-lived 240Pu, for example, unless you want your bomb to melt before it goes off.
Weapons grade plutonium is typically produced using a short fuel cycle (~3 months) that minimizes 240Pu build-up. Power reactors use a much longer fuel cycle, which means that some kind of 239Pu enrichment would be required to turn their spent fuel into something bomb-worthy.
That said, the answer to the question "could terrorists build nuclear bombs" is clearly, "Yes."
It's like suggesting a nuclear plant somewhere in southern California is actually in Los Angeles, which is just plain dumb.
Err... the San Onofre plant is less than 50 km from LA. Not quite in the city, but not so far off, either, and close enough that a serious accident would have a very large effect on the city itself.
I tried to get VOIP via Vonage in Canada, and based on my experience would stay well clear of them. Their technology worked extremely well, but they neglected to mention that while they provide service in my area code, they do not service my local calling area. So even though I was told I could keep my old phone number, I was actually assigned a number that is long distance from here.
They promised a full refund, but then tried to ding me for a cancellation fee.
So based on that experience, I'd recommend that people go with anyone other than Vonage.
I was referring to "marginal cost", which has to do producing an additional unit---i.e. copies---and it may be considered a "fundamental principle" (I'm not sure if it is, IANAE), but it is clear that "price of a good should exceed its marginal cost."
I am also not an economist, but there is a fundamental economic principle that says the equilibirum market price tends to approach the marginal cost--essentially, profit margins become extremely small over time, as competitors strive for greater efficiency and market share.
In the software market, the marginal cost of production is almost zero (the cost of a download, say $0.01). The market size is variable, but frequently quite small--most programs ship fewer than 100,000 units. Even if the equilibrium profit margin were as high as 10%, the total equilibrium market value of an average program would be 0.01*0.1*100,000, or about $100.
This kind of calculation only applies to a mature feature set with open file formats, and doesn't apply to things like embedded software or areas where liability is a big issue, both of which provide barriers to entry into the market and keep margins much higher. But the shrink-wrap, desktop software market is one where all the money is in the early days of upgrades as the feature set matures, or in markets dominated by complex, proprietary file formats that must be frequently exchanged between users. I'll leave it up to the reader to think of examples.
Another way of putting this is: commodity software is worth nothing, unless you have an effective monopoly and can therefore set prices more-or-less arbitrarily. Patent protection is one way of creating such a monopoly.
The problem with nuclear power isn't the big scary scenarios that the mainstream anti-nuclear community put about. The problem is that economics suck, and probably always will. "Successful" national nuclear power programmes are propped up by artificial means--either direct government investment, or special-needs laws like the insurance liability cap, or both.
Sure, coal plants pump out a lot more garbage into the environment than nuclear plants, but coal plants have two big advantages: relatively small events don't wind up writing-off the whole plant; and you can take the damn things apart and fix them relatively cheaply because they aren't radioactive.
It isn't just "unreasonable regulatory burden" that makes nuclear plants expensive--it is the fact that the available energy density is extremely high, and any departure from equilibrium can result in sufficiently high energy density to result in plastic deformation of components of the core. Once that happens they're hellishly expensive to fix. Even relatively routine maintenance is extremely expensive due to the real safety requirements of doing engineering work in a radioactive environment.
"Inherently safe" design for fission reactors is an interesting area of research, and much progress has been made, but it isn't clear that any of them are really as safe as their designers would like to believe. And again, it isn't the possibility of catastrophic, world-ending melt-down that you need to prevent, but relatively minor excursions that will leave the containment intact but make a mess of the core.
Older designs, such as the CANDU (which has a negative temperature coefficient of reactivity, if memory serves, meaning a temperture spike will damp the reaction down) are already more-or-less "melt-down-proof". But they have also proven to be bloody expensive to maintain--far moreso than coal-fired plants run by the same utility.
These are all reasons I got out of the nuclear engineering business many years ago--the core physics of fission power is such that it is very hard to create reactors that are going to be economic to operate over the lifecycle of the plant.
The universe is not populated by natural kinds--groups of individual entities that "naturally" fall into distinct classes. Apparently because of parsimonious physical laws and fairly deep, narrow ecological optima, we do find lots of things that are pretty similar to each other because their existence is possible only within relatively narrow physical bounds or they are the result of selective processes that don't allow for a lot of latitude.
However, despite a couple of millenia of bad philosophy, we should not be fooled into believing that because most of the time we can group things into convenient categories that that will always be the case, and when we have very few examples of individuals in a given category (the nine planets in the solar system, for example) we can expect that our categorization criteria will undergo radical change as we discover more more-or-less similar individuals. And not only will our categorization criteria change: the categories themselves will change, as we try to find the most epistemologically efficient means of grouping individuals.
"Efficiency" in this case is measured by some combination of keeping the number of classes as small as possible for our purposes, and practical ease of categorization (there's no point in having a very small number of classes if the criteria for membership are practially impossible to apply.)
So asking "What is a planet REALLY?" is the wrong question. The right question is: "What is the most useful classification scheme for these objects, given our purposes as human beings?"
There is environmental caution, economic caution, political caution... Different "cautions" are often diametically opposed.
To say simply that you are "on the side of caution" is to say nothing beyond the fact that you are some kind of conservative. You need to give an argument (which will be a moral argument) as to why you choose one caution over the other.
Graphologists may still be used because the quality of their work is being compared to that of other personality profilers, whose record is considerably less than stellar. Psychics who claim to predict events can easily be discredited when the events don't happen. Phrenology was relatively easy to discredit becuase its practice was extremely quantitative. Graphology does not predict events and is extremely qualitative in technique. It is therefore almost impossible to discredit.
Polygraphs share the same characteristics--because polygraph reading is entirely qualitative, proponents can always claim that poor results are due to poor readers, and that if only they had been the ones doing the reading all would have been well.
I'm a practicing scientist and find philosophy of science both interesting and irritating. It's interesting (and important) because questions like, "What makes a good explanation?" aren't quite part of science but still need to be asked. It's irritating because philosphers are by-and-large such complete and utter wankers.
intense rainstorms and presumably snow storms are a potential indicator of global warming as the oceans evaporate off more water as they warm.
Unfortunately, increased droughts are also a potential indicator of global warming, as are increased rains of fishes.
It goes like this: in hot years, people concerned about global warming make a big deal about how hot it is. In non-hot years, people concerned about global warming make a big deal about the how global warming doesn't necessarily imply that it'll get hotter everywhere, but that weather events will be more extreme.
Now, we know that the global climate is not perfectly stable, and evidence from the past ten thousand years (and from previous inter-glacials) indicates that the inter-glacial climate is probably prone to sudden shifts in global average temperature. We are currently in a "moderate" period, but could easily be flipped into a "hot" period by any combination of factors, including anthropogenic forcing.
But knowing that does not seem to me to be sufficient excuse to indulge outselves with claims that basically all observeable weather patterns consitute evidence that global warming is occurring. They sound far too much like the faux-environmentalist argument that if you see fewer of some species around your home it is evidence they are dying out, and if you see more it is evidence that habitat loss has pushed them out of their native locales.
The weirdest thing about the bubble fusion story is that acetone produces neutrons, but water (even heavy water) does not.
Presumably now that the neutron results in acetone have been replicated, researchers will investigate the role of different molecular properties in neutron production rate. Why there should be a coupling between atomic and nuclear phenomena in this case is far from clear, but it is pretty clear now (although one replication doth not a science make) that some nuclear phenomenon is occuring.
If most of the energy is coming out in neutrons (as appears to be the case) then it may be possible to convert the neutron's energy to heat in a jacket of absorbant material surrounding the liquid, so the liquid itself will remain (relatively) cool. The thermalization length for neutrons is typically ~10 cm, so this would limit the size of the configuration pretty tightly, but it appears that the neutron production rate is dominated by chemistry (which really is very weird) rather than any other aspect of the configuration.
Slashdot Mirror
How many times have you asked someone, "What does your code do to solve this problem?" and got a description of an algorithm which, when you finally get to see the source, does not match the code?
In my case, the answer to that question is, "Lots." I have had it happen in pure science (neutrino physcis), applied science (medical physics) and software development (database programming, data analysis, etc.)
I am painfully aware that my own published descriptions of algorithms have often left out minor details that may be critical in some applications, but that page limits in peer-reviewed journals necessitate. It is not uncommon to get a call from someone doing similar work asking for details about what you've done, how you've done it, and in some cases, asking to look at source code.
In contentious areas of science such requests are not always met with full disclosure, which is a sign that the people involved are no longer doing science. They are doing politics. This happens a lot, and it brings the scientific process to a halt on the question at issue.
In the case at hand, the original authors have done a very poor job of describing what they have done, and an extremely poor job of defending their work. Their refusal to publish their source code for their analysis gives credibility to their critics.
There are certainly legitimate cases where code ought not be published. If a lab has spent many, many years developing a framework for solving a certain type of problem and wants to get the most advantage out of that framework before releasing it, they may reasonably want to limit it's disemination for a while. But those sorts of reason don't apply in this case, and the source should be made available to anyone who wants to reproduce their actual results. That would just be good science.
--Tom
The sphericity of the Earth was not generally accepted until much later than the 1st century C.E. More like the 8th. Even then, the way in which people thought about the Earth was radically different than the modern idea of a spherical Earth:
http://abob.libs.uga.edu/bobk/whiteb03.html
The "doctrine of the antipodes" asserted that even if the Earth was spherical, no humans lived on "the other side" because they'd have their feet in the air, wouldn't be able to observe the descent of Christ at the 2nd coming, etc.
Variants of this doctrine persisted long after the nominal sphericity debate had been settled, and I'd argue that until something like the modern view of a spherical Earth, antipodes and all, was generally accepted, it is not quite correct to claim that it was "generally accepted that the earth is round".
Ancient ideas are alien to our own, and it is easy to impose our modern understanding on the words the ancients used, creating great distortion. So I get to disagree with everyone: in the first millenium C.E. people neither believed that the Earth was flat, nor that the Earth was round in the modern sense. They believed the Earth had a special place in the universe, and their understanding of the shape and geography of the Earth grew out of Church doctrine as much as emprical observation.
--Tom
This article contains a host of issues.
The Dark Matter problem is actually the Dark Matter problemS, plural. Galactic dark matter is only the tip of the iceberg, and can be explained by baryonic matter. Dark matter (or energy) on larger scales is a different kettle of fish. A better heading would have been: "Large scale dynamics of the universe", which would take in the horizon problem, the dark matter problem and the dark energy problem.
The article in this regard is a bit like a software requirements document written by a user: it's in terms of projected solutions rather than actual problems. The actual problem is that we don't understand the large-scale dynamics of the universe. The solution may be anything from exotic particles to weird properties of space to alternative dynamics. We just don't know.
The things about alpha changing and tetra-neutrons are cool, but far more likely to be mistakes than new phenomena.
The stuff about high-energy cosmic rays is probably the most interesting, and in fact there are a wealth of phenomena that have been observed by large detectors such as SNO and Kamiokande over the years that really don't make sense. The possibility of new physics at high energies, or entirely novel particles (magnetic monopoles, for example) is quite real, and some of the anomalies observed in these detectors may be indicators of these things.
--Tom
Some of us still do live in caves. And guess what: there are substantial gender inequalities in modern hunter-gatherer societies. Industrial societies have lower gender inequality than any society that has ever existed, in part because machines take the place of muscle, making the only significant difference between men and women irrelevant.
Women got the vote in the most heavily industrialized parts of the world first, and women's rights in the most heavily industrialized parts of the world are still far in advance of anywhere else.
Women in the industrial world can own property, enter into contracts, marry whom they please. This is not the case in many pre-industrial/developing societies in, for example, sub-Saharan Africa.
--Tom
To put a slight twist on the normal definition, for the home user two-factor is defined as:
1) Something you can loose
2) Something you can forget
And biometrics means:
3) Something that can change.
A surgeon of my acquaintance used to claim that there were only two significant medical advances in the past 200 years: anti-biotics, and sterile technique. To that I would add anesthetic, although that's more of an enabling technology than a cure. The effect of everything else has been tiny in comparison.
Probably the largest "minor" advance was the use of radiation to treat cancer. However, the success rate (60% of patients improve, 40% don't) has not changed significantly since the 50's, when Canadian doctors pioneered cobalt therapy. We've spent oodles of money on improvements since then, and some of those improvements have had a big effect for a small number of people, but the big numbers have not changed significantly.
The reality is that the largest increase in human health world-wide would be achieved by taking 100% of medical research dollars in the developed world and spending them on basic infrastructure--like clean water and sewage treatment--in the developing world.
--Tom
What is the "fundamental rule of honest dealing" you are refering to? I am not aware of any rule that says I should not read a web page written to me, about me, that is trivially available to me.
The issue here is the triviality of availablity: the letters were available using the backspace and enter keys. Harvard made no attempt to restrict access other than by not linking them, which is not security. Many people, myself included, believe that if information about myself is freely accessible then there is nothing wrong with accessing it. An unlinked, unsecured page falls well within my definition of freely accessible--it differs from a paper letter left on someone's desk because it was not left on the website but rather put there deliberately, and should not have been put there if it was not meant to be seen.
The people who are morally culpable here are Harvard, for making information freely accessible to applicants and then punishing them for accessing it.
To forestall more invalid analogies: this is totally unlike walking into someone else's unlocked house, where there is a clear presumption in law and culture that I have no right to do so. This is not about picking up money on the sidewalk that I've seen fall out of someone's pocket and have clear knowledge of ownership. This is not like cheating on my wife: I am not breaking any promises I made (assuming Harvard applicants don't sign any undertaking that they will not attempt to access their admissions status by any means whatsoever.)
This is about information. Information about me. Information that is available to me by hitting the backspace and enter key.
Here's an analogy that might be valid: an applicant calls the admissions office and asks the person who answers the phone, "Am I in?" and the person who answers the phone, who happens to have the applicant's file in front of them, glances at it and says, "Looks like you're rejected."
Then that person tells a friend, who tells a friend, and pretty soon Harvard has 119 people calling them. Most of them get someone who says, "We don't know yet." Some get the dunderhead the first person talked to, who tells them "Looks like you're rejected." No one gets told "You're in."
Now, are these people who are calling the admissions office doing anything unethical? They are doing more work than the people who checked the URL online. How does the fact that Harvard had a poorly trained employee who made information available that should not have been an ethical problem for the people who tried to access that information?
Many of those who accessed it were probably doing the moral equivalent of calling the admissions office: they knew the official date hadn't arrived yet, but they thought they'd just check out this possible alternative and see. After all, if they shouldn't get the information, then Harvard would surely not give it to them.
Right?
--Tom
"When you look at the non-industry sponsored research, it's about three to one--three out of every four papers shows an effect," Lai says. "Then, if you look at the industry-funded research, it's almost opposite--only one out of every four papers shows an effect."
Ever try to get a null result published?
I believe that industry-sponsored research is biased. But simply because research is not industry-sponsored does not mean it is not biased.
In particular, what a scientist wants to see in an experiment is a positive effect, an non-null result. I've seen people (in genetics, as it happens) do terrible things to their data to get a non-null result, and carefully massage the statistics to make a result that deviates from the null hypothesis by a miniscule amount look significant.
Why?
Because it's a hell of a lot easier (to say nothing of more personally satisifying) to do all this work, kill all these rats, and at the end of the day be able to say something more interesting than, "Nothing to see here, move along..."
--Tom
The natural move from my point of view is to look at mu-N interactions, where a muon blows apart a nucleus in the target material, producing a shower of excited nuclear fragments and neutrons. Heavy materials such as plutonium will have a much different cascade signature than relatively light things like iron, so it may be possible to develop a quite specific finger-printing mechanism that would be hard to work around. With a muon detector on top to act as a trigger, and some combination of gamma and neutron detectors nearby, this is might be able to both speed up processing and improve accuracy dramatically.
Of course, terrorists could always fall back to the obvious plan B: smuggling the weapon in hidden in a bale of marijuana.
--Tom
They found that just about everyone could, on a small but repeatable level, affect the output of a random number generator just by concentrating on it. (The implications of that, if true, are staggering enough alone)
Extremely careful analysis is required when looking for very small effects in the midst of large masses of data.
See for example: http://quasar.as.utexas.edu/papers/reg.pdf
Frequentist analysis breaks down in a variety of circumstances, and Bayesian analysis must be used instead. The most familiar case where frequentist analysis breaks down is when there are a very small number (or just one) event(s). But it also breaks down in these large datasets when one goes hunting for very small probability events.
Looked at informally, what is more probable: that humans have a small but significant ability to alter events by thinking about them (that evolution has somehow missed out on improving on) or that the experiments and analysis are somehow flawed? Naively, the latter hypothesis is more plausible, and the paper linked above demonstrates this to be the case.
--Tom
In that case you have a specification! In the form of your static prototype.
A prototype and a specification do not contain the same information. A prototype consists of a single concrete instance of the thing a specification describes. It contains more information than the specification in some respects (the concrete design choices the implementor has made to fill in the gaps the specification is silent on) but more importantly it is also missing information that is absolutely required in a specification.
For example: What are the tolerances on particular processes? What are the scaling and responsiveness requirements? And so on.
Nothing that does not capture this kind of information is a specification, and so far as I know only a document can capture this kind of information.
Don't get me wrong: prototypes are useful. But they are not specifications, because they do not contain the all of the information that a specification must contain.
--Tom
The key is he has top notch technical & marketing leaders on his team that guide the overall technical direction.
No, the key is that he listens to them and trusts them when they advise him on things he does not understand. Lots of companies have good technical advisors. Very few pay any attention to them.
--Tom
De jure Canada has one-tier systems. De facto we are multi-tier, and becoming moreso. In education, kids who go to York or Carleton don't get the same quality of education as kids who go to Queen's or even U of T. And for the rich there are much better health care options available, including private clinics in the U.S.
The political mythology of a single-tier system has helped to maintain the quality of that system until the past few years, but the reality is that Canadians have more choices available than they are often willing to admit to. Nor are our taxes that much higher, and they are likely to get lower due to our more fiscally responsible governments.
Canadian high schools are better than their U.S. counter-parts, but it's my impression that local alternatives are a larger force here, including mixed schools where a number of different educational philosphies or foci are represented by different organizations within a single physical building. This diversity, both within and without the public system, is one of the things that is keeping our public schools from completely failing in their charge.
--Tom
I saw a talk by a patent agent/IP lawyer that was basically FUD about FOSS, and one of his big points was, "There are more than forty different open source licenses out there!" He didn't mention that there are thousands of different closed source licenses, and somehow the business community has not ground to a halt while trying to deal with their complexity and diversity.
I think that might have been because people like him make their money out of all that complexity, and so the simplicity and directness of most open source licenses are a profound threat to their income. We are about as likely to see a standard set of closed source licenses as we are to see a national standard building code, and for the same reason: there are powerful economic interests opposed to them.
--Tom
30 kg of plutonium is more than enough. There are reprocessing issues even with plutonium, though--you need to get rid of the most of the relatively short-lived 240Pu, for example, unless you want your bomb to melt before it goes off.
Weapons grade plutonium is typically produced using a short fuel cycle (~3 months) that minimizes 240Pu build-up. Power reactors use a much longer fuel cycle, which means that some kind of 239Pu enrichment would be required to turn their spent fuel into something bomb-worthy.
That said, the answer to the question "could terrorists build nuclear bombs" is clearly, "Yes."
--Tom
It's like suggesting a nuclear plant somewhere in southern California is actually in Los Angeles, which is just plain dumb.
Err... the San Onofre plant is less than 50 km from LA. Not quite in the city, but not so far off, either, and close enough that a serious accident would have a very large effect on the city itself.
--Tom
I tried to get VOIP via Vonage in Canada, and based on my experience would stay well clear of them. Their technology worked extremely well, but they neglected to mention that while they provide service in my area code, they do not service my local calling area. So even though I was told I could keep my old phone number, I was actually assigned a number that is long distance from here.
They promised a full refund, but then tried to ding me for a cancellation fee.
So based on that experience, I'd recommend that people go with anyone other than Vonage.
--Tom
I was referring to "marginal cost", which has to do producing an additional unit---i.e. copies---and it may be considered a "fundamental principle" (I'm not sure if it is, IANAE), but it is clear that "price of a good should exceed its marginal cost."
I am also not an economist, but there is a fundamental economic principle that says the equilibirum market price tends to approach the marginal cost--essentially, profit margins become extremely small over time, as competitors strive for greater efficiency and market share.
In the software market, the marginal cost of production is almost zero (the cost of a download, say $0.01). The market size is variable, but frequently quite small--most programs ship fewer than 100,000 units. Even if the equilibrium profit margin were as high as 10%, the total equilibrium market value of an average program would be 0.01*0.1*100,000, or about $100.
This kind of calculation only applies to a mature feature set with open file formats, and doesn't apply to things like embedded software or areas where liability is a big issue, both of which provide barriers to entry into the market and keep margins much higher. But the shrink-wrap, desktop software market is one where all the money is in the early days of upgrades as the feature set matures, or in markets dominated by complex, proprietary file formats that must be frequently exchanged between users. I'll leave it up to the reader to think of examples.
Another way of putting this is: commodity software is worth nothing, unless you have an effective monopoly and can therefore set prices more-or-less arbitrarily. Patent protection is one way of creating such a monopoly.
--Tom
The problem with nuclear power isn't the big scary scenarios that the mainstream anti-nuclear community put about. The problem is that economics suck, and probably always will. "Successful" national nuclear power programmes are propped up by artificial means--either direct government investment, or special-needs laws like the insurance liability cap, or both.
Sure, coal plants pump out a lot more garbage into the environment than nuclear plants, but coal plants have two big advantages: relatively small events don't wind up writing-off the whole plant; and you can take the damn things apart and fix them relatively cheaply because they aren't radioactive.
It isn't just "unreasonable regulatory burden" that makes nuclear plants expensive--it is the fact that the available energy density is extremely high, and any departure from equilibrium can result in sufficiently high energy density to result in plastic deformation of components of the core. Once that happens they're hellishly expensive to fix. Even relatively routine maintenance is extremely expensive due to the real safety requirements of doing engineering work in a radioactive environment.
"Inherently safe" design for fission reactors is an interesting area of research, and much progress has been made, but it isn't clear that any of them are really as safe as their designers would like to believe. And again, it isn't the possibility of catastrophic, world-ending melt-down that you need to prevent, but relatively minor excursions that will leave the containment intact but make a mess of the core.
Older designs, such as the CANDU (which has a negative temperature coefficient of reactivity, if memory serves, meaning a temperture spike will damp the reaction down) are already more-or-less "melt-down-proof". But they have also proven to be bloody expensive to maintain--far moreso than coal-fired plants run by the same utility.
These are all reasons I got out of the nuclear engineering business many years ago--the core physics of fission power is such that it is very hard to create reactors that are going to be economic to operate over the lifecycle of the plant.
--Tom
In borderline cases concepts proliferate.
The universe is not populated by natural kinds--groups of individual entities that "naturally" fall into distinct classes. Apparently because of parsimonious physical laws and fairly deep, narrow ecological optima, we do find lots of things that are pretty similar to each other because their existence is possible only within relatively narrow physical bounds or they are the result of selective processes that don't allow for a lot of latitude.
However, despite a couple of millenia of bad philosophy, we should not be fooled into believing that because most of the time we can group things into convenient categories that that will always be the case, and when we have very few examples of individuals in a given category (the nine planets in the solar system, for example) we can expect that our categorization criteria will undergo radical change as we discover more more-or-less similar individuals. And not only will our categorization criteria change: the categories themselves will change, as we try to find the most epistemologically efficient means of grouping individuals.
"Efficiency" in this case is measured by some combination of keeping the number of classes as small as possible for our purposes, and practical ease of categorization (there's no point in having a very small number of classes if the criteria for membership are practially impossible to apply.)
So asking "What is a planet REALLY?" is the wrong question. The right question is: "What is the most useful classification scheme for these objects, given our purposes as human beings?"
--Tom
I guess I prefer to err on the side of caution.
Which caution?
There is environmental caution, economic caution, political caution... Different "cautions" are often diametically opposed.
To say simply that you are "on the side of caution" is to say nothing beyond the fact that you are some kind of conservative. You need to give an argument (which will be a moral argument) as to why you choose one caution over the other.
--Tom
Graphologists may still be used because the quality of their work is being compared to that of other personality profilers, whose record is considerably less than stellar. Psychics who claim to predict events can easily be discredited when the events don't happen. Phrenology was relatively easy to discredit becuase its practice was extremely quantitative. Graphology does not predict events and is extremely qualitative in technique. It is therefore almost impossible to discredit.
Polygraphs share the same characteristics--because polygraph reading is entirely qualitative, proponents can always claim that poor results are due to poor readers, and that if only they had been the ones doing the reading all would have been well.
--Tom
I'm a practicing scientist and find philosophy of science both interesting and irritating. It's interesting (and important) because questions like, "What makes a good explanation?" aren't quite part of science but still need to be asked. It's irritating because philosphers are by-and-large such complete and utter wankers.
--Tom
intense rainstorms and presumably snow storms are a potential indicator of global warming as the oceans evaporate off more water as they warm.
Unfortunately, increased droughts are also a potential indicator of global warming, as are increased rains of fishes.
It goes like this: in hot years, people concerned about global warming make a big deal about how hot it is. In non-hot years, people concerned about global warming make a big deal about the how global warming doesn't necessarily imply that it'll get hotter everywhere, but that weather events will be more extreme.
Now, we know that the global climate is not perfectly stable, and evidence from the past ten thousand years (and from previous inter-glacials) indicates that the inter-glacial climate is probably prone to sudden shifts in global average temperature. We are currently in a "moderate" period, but could easily be flipped into a "hot" period by any combination of factors, including anthropogenic forcing.
But knowing that does not seem to me to be sufficient excuse to indulge outselves with claims that basically all observeable weather patterns consitute evidence that global warming is occurring. They sound far too much like the faux-environmentalist argument that if you see fewer of some species around your home it is evidence they are dying out, and if you see more it is evidence that habitat loss has pushed them out of their native locales.
--Tom
The weirdest thing about the bubble fusion story is that acetone produces neutrons, but water (even heavy water) does not.
Presumably now that the neutron results in acetone have been replicated, researchers will investigate the role of different molecular properties in neutron production rate. Why there should be a coupling between atomic and nuclear phenomena in this case is far from clear, but it is pretty clear now (although one replication doth not a science make) that some nuclear phenomenon is occuring.
If most of the energy is coming out in neutrons (as appears to be the case) then it may be possible to convert the neutron's energy to heat in a jacket of absorbant material surrounding the liquid, so the liquid itself will remain (relatively) cool. The thermalization length for neutrons is typically ~10 cm, so this would limit the size of the configuration pretty tightly, but it appears that the neutron production rate is dominated by chemistry (which really is very weird) rather than any other aspect of the configuration.
--Tom