On the plus side, Dayan and Abbott is actually a graduate-level text, but you're utterly right. To keep up with just the unsupervised learning methods that it covers requires at least a stats course, a linear algebra course, and a couple of calculus courses. To make things worse the newest edition is 8 years old, which is a significant portion of the lifetime of the modern cogsci field.
You might appreciate this blog, which is at least about availability and help might get you more well-read, but even disregarding the educational gap, the field is the intersection of ML and medicine, two areas that are both extremely high-pressure and high-prestige. It will probably be the last discipline that breaks out of the cathedral and finds the bazaar.
It's important to not over-generalize when talking about genetically-modified foods. Monsanto uses a particularly unsafe technique to do of its lot of engineering where they simply bombard plants with mutagens until they get what they want; the normal strategy that comes to mind (splicing genes selectively) has a very low chance of causing human health problems. By contrast, mutagenic treatment just Fucks Shit Up indiscriminately.
Also, you're more than likely already a guinea pig, so how would you do controls?
I wouldn't go so far as to assume that the FDA is completely overrun. A little under 50% of drugs fail FDA approval on their first application. FDA rejection is costly, and companies have been increasingly been aggressive about doing their own testing first in order to make sure that they don't languish forever in a nightmarish backlog like the one that the USPTO suffers from. I used to know someone who had exactly the sort of near-executive-level pharmaceutical responsibility; as far as I could tell, a lot of the collaboration between FDA people and companies is actually about trying to expedite testing and safety.
On top of that, you have competitive pressures. Nothing is better for a company if they can discover that their competitors have cheated regulations or produced an unsafe product; the battlefield is aggressive and collaborations usually end in backstabbing. If you can produce evidence that another company lied to the FDA or that their products pose a health risk, it can potentially destroy that company. This is one case where a competitive market can be a positive force if the rules are set up right.
That all being said, the FDA doeshave corruption issues. The Wikipedia article on on regulatory capture lists some much more perverse cases, though, like how the agency responsible for cleaning up after oil spills was renamed and then restructured into oblivion in the days following the Deepwater Horizon spill.
Like it or not, even big companies are innocent until proven guilty. Pending FDA approval, anyway.
In this case it looks like the researchers were out for blood and let their dislike for Monsanto get in the way of doing the science properly—not only did they use cancer-prone rats like it says in the summary, but they didn't do enough replicates to determine if the results were actually statistically significant: the control group definitely got fewer tumours, but given the unreliability of the rat breed's tumour-forming rate it's hard to say that it wasn't just a coincidence. (And using a cancer-prone rat isn't exactly realistic to begin with; tumours grow faster whenever they get cheap and easy nutrients.)
The paper was under close scrutiny immediately when it was published, and not just from Elsevier or Monsanto.
Rest assured, I thought it was hilarious. But, sadly, it really was just a typo and Microsoft does not really outsource all of their programming. (What would they do with all the H1Bs that they keep lobbying for?)
In this case the people being monitored are leaders of radical Islamic groups in Pakistan. We're not talking about the Weather Underground here; there are completely different cultural norms at play.
I would be completely happy with that assessment if it weren't for the fact that many neoreactionaries are former libertarians (it's in TFA.)
The problem is—and I don't think this is much of a stretch—the mainstream left-wing perspective is that libertarian policies create huge economic disparities, and that libertarians are either wilfully ignorant of the inevitability of lobbying and monopoly-protecting regulations, or actually welcome such a situation and are lying about it. To such a perspective, neoreactionaries are simply being honest about their misanthropy.
I'm not saying the two movements are equatable, just that it makes libertarianism looks bad, in the same way radical Christians make Christianity look bad—many of the core beliefs are different, but only to supplicate those burdened by the moral demands of the mainstream belief.
The concept of Social Darwinism as generally discussed only uses the "value generation = success" premise as an excuse; I'm afraid you've misunderstood that part. The actual claims were that different ethnicities (wildly overgeneralized as "races") have different intellectual potentials based on strictly genetic traits, which remains unproven. Data based on IQ, especially older IQ tests, is almost certainly an artefact of deep-seated cultural differences. There's a long and uncomfortable history of this.
If "humanist" is too vague for you, let me say "empathetic rationalist" instead.
Oh, but look what they said: "Each dollar goes further towards improving standard of living for the average person in an aristocratic system than in a Democratic one."
I would be hard-pressed to invent a better parody of libertarianism... or a better way of discrediting it.
The article goes into much more agonizing detail—they literally want to bring back social Darwinism and believe in IQ as both valid and essential to determining a person's worth. I never thought I'd see something that makes Ayn Rand sound like a humanist.
A loud clatter of gunk music flooded through the Heart of Gold cabin as Zaphod searched the sub-etha
radio wavebands for news of himself. The machine was rather difficult to operate. For years radios had been
operated by means of pressing buttons and turning dials; then as the technology became more sophisticated
the controls were made touch-sensitive—you merely had to brush the panels with your fingers; now all you
had to do was wave your hand in the general direction of the components and hope. It saved a lot of
muscular expenditure of course, but meant that you had to sit infuriatingly still if you wanted to keep listening
to the same programme.
Hold your horses: that shit takes significant amounts of time. Unless there were snow-topped mountains in the Yucatan, there was no ice to acclimatize to. It's easy to adapt to periodic ice—even if every single bacterium gets wiped out one year, new bacteria can still enter from the outside environment each year until an antifreeze protein is developed and a foothold established.
The more extreme the new conditions, the harder life has to work to adapt. As it is, there are only a handful of bacteria (examples) that can withstand the conditions typical of Mars, and they had a great deal of time to practice. And that's not counting the expected dose of DNA-shredding radiation, which is bearable by even fewer critters. (The tardigrade being one hilarious exception—but they go into hibernation mode when in space and would never wake up on Europa's surface.)
Interestingly, the Atacama Desert in Bolivia seems to have terrain high enough for bacteria to evolve considerable radioresistance, amongst other things. If the Yucatan had enough height at the time, there's a tiny chance it was a similar biome, but in general the Cretaceous period is known to have been very humid, so this seems unlikely.
I was thinking that might be the trick, although the region is still not very suitable for honing and maintaining cold-weather survival skills. Here's hoping the crater was pretty deep.
It's pretty problematic that the impact in question happened in Mexico. The Yucatan isn't exactly a haven of extremophiles—you wouldn't expect to find anything that can maintain a biosphere without a good light source, and they're definitely not well-adapted to the sulphur and magnesium contamination that Europa appears to have. Unfortunately the best places to find organisms with a chance of surviving in this kind of environment are at the bottom of the ocean, which is a particularly bad target for producing ejecta. Caves are also a possibility, and since Mexico has no shortage of them, they might be a potential avenue... but who knows if there were any decent ones in the Yucatan at the time.
No, it was a Russian contractor that serviced many of those facilities. The infections appeared to have been caused by him. We talked about this in a previous Schneier post on Stuxnet. Whether he was paid to do it or his equipment subverted I don't know.
Woah, woah, I'm not implicating the Russian government. I'm just saying that Stuxnet infections are common in Russia. Everyone believes the ISS infection was an accident.
Stuxnet was delivered to Iran by slipping it onto the equipment of the Russian contractors building the nuclear plant.
Gauss was discovered in Lebanon and appears to have been built with the same toolkit, not reverse-engineered, suggesting the Israelis were responsible for its release. The other known variant, Flame, is also not found anywhere near Russia, and was also cut from the same cloth and targeted at Iran.
And, moreover, Kaspersky said it was Stuxnet, with (I'm pretty sure but don't have the time to watch the whole video) implications that it may've come by accident from an attack targeted at a Russian nuclear plant.
So... still pretty unlikely the ISS has any vulnerable systems.
Slashdot Mirror
On the plus side, Dayan and Abbott is actually a graduate-level text, but you're utterly right. To keep up with just the unsupervised learning methods that it covers requires at least a stats course, a linear algebra course, and a couple of calculus courses. To make things worse the newest edition is 8 years old, which is a significant portion of the lifetime of the modern cogsci field.
You might appreciate this blog, which is at least about availability and help might get you more well-read, but even disregarding the educational gap, the field is the intersection of ML and medicine, two areas that are both extremely high-pressure and high-prestige. It will probably be the last discipline that breaks out of the cathedral and finds the bazaar.
It's important to not over-generalize when talking about genetically-modified foods. Monsanto uses a particularly unsafe technique to do of its lot of engineering where they simply bombard plants with mutagens until they get what they want; the normal strategy that comes to mind (splicing genes selectively) has a very low chance of causing human health problems. By contrast, mutagenic treatment just Fucks Shit Up indiscriminately.
Also, you're more than likely already a guinea pig, so how would you do controls?
I wouldn't go so far as to assume that the FDA is completely overrun. A little under 50% of drugs fail FDA approval on their first application. FDA rejection is costly, and companies have been increasingly been aggressive about doing their own testing first in order to make sure that they don't languish forever in a nightmarish backlog like the one that the USPTO suffers from. I used to know someone who had exactly the sort of near-executive-level pharmaceutical responsibility; as far as I could tell, a lot of the collaboration between FDA people and companies is actually about trying to expedite testing and safety.
On top of that, you have competitive pressures. Nothing is better for a company if they can discover that their competitors have cheated regulations or produced an unsafe product; the battlefield is aggressive and collaborations usually end in backstabbing. If you can produce evidence that another company lied to the FDA or that their products pose a health risk, it can potentially destroy that company. This is one case where a competitive market can be a positive force if the rules are set up right.
That all being said, the FDA does have corruption issues. The Wikipedia article on on regulatory capture lists some much more perverse cases, though, like how the agency responsible for cleaning up after oil spills was renamed and then restructured into oblivion in the days following the Deepwater Horizon spill.
Like it or not, even big companies are innocent until proven guilty. Pending FDA approval, anyway.
In this case it looks like the researchers were out for blood and let their dislike for Monsanto get in the way of doing the science properly—not only did they use cancer-prone rats like it says in the summary, but they didn't do enough replicates to determine if the results were actually statistically significant: the control group definitely got fewer tumours, but given the unreliability of the rat breed's tumour-forming rate it's hard to say that it wasn't just a coincidence. (And using a cancer-prone rat isn't exactly realistic to begin with; tumours grow faster whenever they get cheap and easy nutrients.)
The paper was under close scrutiny immediately when it was published, and not just from Elsevier or Monsanto.
That... is bizarre. That's probably like a hundredth of the market demand, if not less.
That is mindblowingly sad. :( I wanted one, too!
THE STORY SHALL BE REPOSTED UNTIL THE CRYING CEASES.
Think of it as less of a dupe and more of a commandment to buy one.
Rest assured, I thought it was hilarious. But, sadly, it really was just a typo and Microsoft does not really outsource all of their programming. (What would they do with all the H1Bs that they keep lobbying for?)
I'm not; you're just that unfunny, as evidenced by the moderation on your post. Sorry, champ.
The link goes to a summary that says $900 million. It was a simple typo. Good luck on your next round of RTFA Roulette!
Whatever the reality, it's still extremely rude and immoral expose.
In this case the people being monitored are leaders of radical Islamic groups in Pakistan. We're not talking about the Weather Underground here; there are completely different cultural norms at play.
I'm going to take that as a sign of agreement. Have fun ruining your country!
I would be completely happy with that assessment if it weren't for the fact that many neoreactionaries are former libertarians (it's in TFA.)
The problem is—and I don't think this is much of a stretch—the mainstream left-wing perspective is that libertarian policies create huge economic disparities, and that libertarians are either wilfully ignorant of the inevitability of lobbying and monopoly-protecting regulations, or actually welcome such a situation and are lying about it. To such a perspective, neoreactionaries are simply being honest about their misanthropy.
I'm not saying the two movements are equatable, just that it makes libertarianism looks bad, in the same way radical Christians make Christianity look bad—many of the core beliefs are different, but only to supplicate those burdened by the moral demands of the mainstream belief.
The concept of Social Darwinism as generally discussed only uses the "value generation = success" premise as an excuse; I'm afraid you've misunderstood that part. The actual claims were that different ethnicities (wildly overgeneralized as "races") have different intellectual potentials based on strictly genetic traits, which remains unproven. Data based on IQ, especially older IQ tests, is almost certainly an artefact of deep-seated cultural differences. There's a long and uncomfortable history of this.
If "humanist" is too vague for you, let me say "empathetic rationalist" instead.
Oh, but look what they said: "Each dollar goes further towards improving standard of living for the average person in an aristocratic system than in a Democratic one."
I would be hard-pressed to invent a better parody of libertarianism... or a better way of discrediting it.
The article goes into much more agonizing detail—they literally want to bring back social Darwinism and believe in IQ as both valid and essential to determining a person's worth. I never thought I'd see something that makes Ayn Rand sound like a humanist.
A loud clatter of gunk music flooded through the Heart of Gold cabin as Zaphod searched the sub-etha radio wavebands for news of himself. The machine was rather difficult to operate. For years radios had been operated by means of pressing buttons and turning dials; then as the technology became more sophisticated the controls were made touch-sensitive—you merely had to brush the panels with your fingers; now all you had to do was wave your hand in the general direction of the components and hope. It saved a lot of muscular expenditure of course, but meant that you had to sit infuriatingly still if you wanted to keep listening to the same programme.
The other day I got ripped off during a TF2 item trade with Gabe Newell!
Hold your horses: that shit takes significant amounts of time. Unless there were snow-topped mountains in the Yucatan, there was no ice to acclimatize to. It's easy to adapt to periodic ice—even if every single bacterium gets wiped out one year, new bacteria can still enter from the outside environment each year until an antifreeze protein is developed and a foothold established.
The more extreme the new conditions, the harder life has to work to adapt. As it is, there are only a handful of bacteria (examples) that can withstand the conditions typical of Mars, and they had a great deal of time to practice. And that's not counting the expected dose of DNA-shredding radiation, which is bearable by even fewer critters. (The tardigrade being one hilarious exception—but they go into hibernation mode when in space and would never wake up on Europa's surface.)
Interestingly, the Atacama Desert in Bolivia seems to have terrain high enough for bacteria to evolve considerable radioresistance, amongst other things. If the Yucatan had enough height at the time, there's a tiny chance it was a similar biome, but in general the Cretaceous period is known to have been very humid, so this seems unlikely.
I was thinking that might be the trick, although the region is still not very suitable for honing and maintaining cold-weather survival skills. Here's hoping the crater was pretty deep.
It's pretty problematic that the impact in question happened in Mexico. The Yucatan isn't exactly a haven of extremophiles—you wouldn't expect to find anything that can maintain a biosphere without a good light source, and they're definitely not well-adapted to the sulphur and magnesium contamination that Europa appears to have. Unfortunately the best places to find organisms with a chance of surviving in this kind of environment are at the bottom of the ocean, which is a particularly bad target for producing ejecta. Caves are also a possibility, and since Mexico has no shortage of them, they might be a potential avenue... but who knows if there were any decent ones in the Yucatan at the time.
From a comment on Bruce Schneier's blog:
No, it was a Russian contractor that serviced many of those facilities. The infections appeared to have been caused by him. We talked about this in a previous Schneier post on Stuxnet. Whether he was paid to do it or his equipment subverted I don't know.
Woah, woah, I'm not implicating the Russian government. I'm just saying that Stuxnet infections are common in Russia. Everyone believes the ISS infection was an accident.
Stuxnet was delivered to Iran by slipping it onto the equipment of the Russian contractors building the nuclear plant.
Gauss was discovered in Lebanon and appears to have been built with the same toolkit, not reverse-engineered, suggesting the Israelis were responsible for its release. The other known variant, Flame, is also not found anywhere near Russia, and was also cut from the same cloth and targeted at Iran.
And, moreover, Kaspersky said it was Stuxnet, with (I'm pretty sure but don't have the time to watch the whole video) implications that it may've come by accident from an attack targeted at a Russian nuclear plant.
So... still pretty unlikely the ISS has any vulnerable systems.