So I went away and actually did the (literature) research on this. It turns out the answer is a little more complicated than fat suppresses appetite and sugar doesn't.
Basically, what little we understand of appetite control points to three hormones: leptin, insulin and ghrelin (great name). Oversimplifying massively, insulin tells you not to eat when your blood sugar is high and to be hungry when your blood sugar is low; leptin tells you not to eat if your stored fat is high and to be hungry when it is too low; and ghrelin tells you not to eat if your stomach is full and to be hungry if your stomach is empty.
However, to complicate things further, firstly these hormones interact with each other such that, for instance, if your stored fat is high, you will be less sensitive to low blood sugar or an empty stomach. However, you also adjust your long term sensitivity to these hormone levels to make the average level become the baseline, so if you're fat for a sustained period, your sensitivity to leptin will decrease and therefore low blood sugar levels or an empty stomach will revert to causing normal levels of hunger.
What this means is that sustained consumption of high levels of either fats or sugars will lead to increased hunger levels and therefore obesity. This sounds like a no brainer, but note that it is a cumulative effect, in that increased resistance to these appetite suppressing hormones makes it more likely that your body will fool you into overeating, which will make it more likely that your resistance will increase.
Finally, to link it all back to the article, the mechanisms controlling our sensitivity to these hormones is still poorly understood and known to be affected by such things as sleep patterns, stress and illness. It's therefore not too much of a stretch to say that environmental factors are causing a long term decrease in sensitivity. It would been nice to have a bit more of this detailed information in the article though and a little less 'it's something to do with leptin but we're not sure what'.
Seriously, did you RTFA where the researcher himself described the algorithm as being based on PageRank
“We first experimented with our own ideas on network algorithms until we realized that what we needed existed already with the PageRank algorithm, so why reinvent the wheel?” Winter recalled. “Our PageRank-based algorithm singles out proteins in the cancer cells that seem to either promote or suppress disease progression,” Winter said.
You know (bear with me on this), one thing that really annoys me on the internet is when someone spends considerable time and effort putting together a humorous photoshop/blog post/top 10, and the next day I see it, completely uncredited in a national newspaper. Some journalist has stolen it...just because it's on the internet...and stuff on the internet is like, free, right?
Much as I hate to admit it, I think on this occasion Michael Lyton has a point (dammit, I don't like what he says but I have to defend his right to say it). In the real world, no one would seriously contemplate reprinting the contents of a book they borrowed from the library and passing it off as their own, and no one would seriously contemplate walking into their local record store and walking out with anything that caught their eye just because they 'wouldn't have bought it anyway if they'd had to pay full price'.
Thing is, I also buy into the argument that illegal copying actually promotes music sales. Hell, I copied enough albums from my friends when I was a kid to know that I still bought a lot of albums. But don't try to con me that what I wasn't doing wasn't stealing (i.e. taking without permission). It's stealing when a journalist tries to pass off my website as his own work, it's stealing when I copy an album that I never wanted to listen to but my friend says I might quite like, and it's stealing when I download the latest star trek movie because I can't be bothered to pay for it at the cinema and after all, it's bound to be shown on free television at some point anyway.
So let's reboot this discussion. All illegal downloading is theft. Full stop. The more interesting question, is it theft like stealing a pen from work, or is it theft like stealing a car. And if it's theft like stealing a pen, then why is so much more like stealing a car when somebody does it to me.
Well, I guess we can rely on ego and patronage at roughly the same time as we can rely on ego and patronage to encourage people to keep on working at any other job. For instance, we could rely on computer programmers to come into work for the honour of working for the software company and the prestige of producing really cool software.
But wait, I hear you say - some of us need to eat. Aha, so that's where the patronage comes in. In return for providing services to the software company (or whichever job you happen to be in) the software company provides sufficient money for its employees to work, travel, clothe themselves etc.
This is great, so wherever somebody needs a service out of somebody, they should pay them some money. That way everyone gets to eat.
But what if the group of people who want the service is a very large, unconnected group of people. They don't have a large pot of money with which to pay for the service they require. Nor do they have any kind of forum in which to agree to club together to pay for the service. Theoretically, all of the people who wish to benefit from the service could contribute part of the cost. In fact, that's a great idea - why doesn't everyone who want to benefit from the service (let's assume the service is the production of some music they'd like to listen to) pay a fixed cost in order to access the results of that service, and in return they are given the means to access that service (say a licensed copy of the music).
Of course, this system breaks down if freeloaders enter the system and start accessing the services without paying for them. You could rely on sufficient honest people in the system to appropriately pay the service provider enough money for them to eat (they'll probably see about 50p from each album sale so that requires approximately 20,000 album sales a year to get the same salary as could be achieved by working as a cleaner in McDonalds), but most people are only honest when the rules are clearly defined. So we need a clear rule about when it is appropriate to contribute to the service provider - in this case that would be the mechanical copyright law.
So, copyright law is an extremely fair way as stating that if someone wants to enjoy the fruits of someone's labours, they need to pay that person. It breaks down where large corporation acquire copyright and then overstep the mark in terms of collecting on their copyright (e.g. demanding money because someone possessed the means to copy copyrighted material, but where there is no evidence that they did so), but where someone has a reasonable expectation of payment for their services, it is difficult to come up with a fairer mechanism for ensuring that payment.
Uh no. "Image sensors", like eyes (incidentally, how many other image sensors do _you_ have) do an awfully large amount of preprocessing before the signal is sent to the brain. The signal is only similar to other nerve signals in its transmission media and encoding principles. The data itself is quite different (c.f. data transmission and voice transmission over a CDMA network).
You're right however, when you say that this still needs to be translated by the brain. You're assumption appears to be, however, that if you can't see at birth then you'll never be able to see. In truth, no more than approx 30% of congenital cases of blindness are cortical in origin, the vast majority being related to developmental problems in one or both eyes (microphthalmos, anophthalmos etc.) In cases involving cortical disfunction, then clearly this technique is unlikely to work, but where there is no cortical abnormality the brain remains plastic enough to learn to see no matter how late the signals start arriving.
One landmark case involved a patient who had cataracts from birth being given surgery to remove these cataracts in their 30s. The patient was initially able to see only colours and lights. After some time, the patient began to recognise shapes and eventually was able to process visual stimuli to a similar degree to a non-handicapped individual. Unfortunately, the patient did not adapt well to being able to see, finding it distracting and unpleasant. They also felt exluded from their community of visually impaired friends. The issue was not that the brain was unable to adapt functionally to visual stimuli, but the patient was behaviourally adapted to life without vision.
The upshot of this is that restoration of vision to the congenitally blind is not new and is limited by the brains adaptability. What is new is the range of conditions that could be treated by an advance such as this (for the non-classically trained amongst you, anopthalamos means being born without any eyes - try treating that with drops).
I find if you mock out everything the unit tests run great. 100% pass rate every time. The coverage is a little low though.
Mocking out database access is fine when all you care about is using the resulting data (e.g. when testing post processing of the data). However, when testing the actual database interaction (e.g. SQL generation, SQL driven data processing, data access layer etc.) there is no substitute for backing on to a real database to uncover real errors.
Then all you need to worry about is cross database support, cross platform support, pre-population of the database, roll-back, comparison against expected results...
Slashdot Mirror
So I went away and actually did the (literature) research on this. It turns out the answer is a little more complicated than fat suppresses appetite and sugar doesn't.
Basically, what little we understand of appetite control points to three hormones: leptin, insulin and ghrelin (great name). Oversimplifying massively, insulin tells you not to eat when your blood sugar is high and to be hungry when your blood sugar is low; leptin tells you not to eat if your stored fat is high and to be hungry when it is too low; and ghrelin tells you not to eat if your stomach is full and to be hungry if your stomach is empty.
However, to complicate things further, firstly these hormones interact with each other such that, for instance, if your stored fat is high, you will be less sensitive to low blood sugar or an empty stomach. However, you also adjust your long term sensitivity to these hormone levels to make the average level become the baseline, so if you're fat for a sustained period, your sensitivity to leptin will decrease and therefore low blood sugar levels or an empty stomach will revert to causing normal levels of hunger.
What this means is that sustained consumption of high levels of either fats or sugars will lead to increased hunger levels and therefore obesity. This sounds like a no brainer, but note that it is a cumulative effect, in that increased resistance to these appetite suppressing hormones makes it more likely that your body will fool you into overeating, which will make it more likely that your resistance will increase.
Finally, to link it all back to the article, the mechanisms controlling our sensitivity to these hormones is still poorly understood and known to be affected by such things as sleep patterns, stress and illness. It's therefore not too much of a stretch to say that environmental factors are causing a long term decrease in sensitivity. It would been nice to have a bit more of this detailed information in the article though and a little less 'it's something to do with leptin but we're not sure what'.
Seriously, did you RTFA where the researcher himself described the algorithm as being based on PageRank
“We first experimented with our own ideas on network algorithms until we realized that what we needed existed already with the PageRank algorithm, so why reinvent the wheel?” Winter recalled.
“Our PageRank-based algorithm singles out proteins in the cancer cells that seem to either promote or suppress disease progression,” Winter said.
How about the abstract of the research paper in question, which specifically mentions Google Page Rank
http://www.ploscompbiol.org/article/info:doi/10.1371/journal.pcbi.1002511
Are you sure it's the journalist who's the retard here?
You know (bear with me on this), one thing that really annoys me on the internet is when someone spends considerable time and effort putting together a humorous photoshop/blog post/top 10, and the next day I see it, completely uncredited in a national newspaper. Some journalist has stolen it...just because it's on the internet...and stuff on the internet is like, free, right?
Much as I hate to admit it, I think on this occasion Michael Lyton has a point (dammit, I don't like what he says but I have to defend his right to say it). In the real world, no one would seriously contemplate reprinting the contents of a book they borrowed from the library and passing it off as their own, and no one would seriously contemplate walking into their local record store and walking out with anything that caught their eye just because they 'wouldn't have bought it anyway if they'd had to pay full price'.
Thing is, I also buy into the argument that illegal copying actually promotes music sales. Hell, I copied enough albums from my friends when I was a kid to know that I still bought a lot of albums. But don't try to con me that what I wasn't doing wasn't stealing (i.e. taking without permission). It's stealing when a journalist tries to pass off my website as his own work, it's stealing when I copy an album that I never wanted to listen to but my friend says I might quite like, and it's stealing when I download the latest star trek movie because I can't be bothered to pay for it at the cinema and after all, it's bound to be shown on free television at some point anyway.
So let's reboot this discussion. All illegal downloading is theft. Full stop. The more interesting question, is it theft like stealing a pen from work, or is it theft like stealing a car. And if it's theft like stealing a pen, then why is so much more like stealing a car when somebody does it to me.
Well, I guess we can rely on ego and patronage at roughly the same time as we can rely on ego and patronage to encourage people to keep on working at any other job. For instance, we could rely on computer programmers to come into work for the honour of working for the software company and the prestige of producing really cool software.
But wait, I hear you say - some of us need to eat. Aha, so that's where the patronage comes in. In return for providing services to the software company (or whichever job you happen to be in) the software company provides sufficient money for its employees to work, travel, clothe themselves etc.
This is great, so wherever somebody needs a service out of somebody, they should pay them some money. That way everyone gets to eat.
But what if the group of people who want the service is a very large, unconnected group of people. They don't have a large pot of money with which to pay for the service they require. Nor do they have any kind of forum in which to agree to club together to pay for the service. Theoretically, all of the people who wish to benefit from the service could contribute part of the cost. In fact, that's a great idea - why doesn't everyone who want to benefit from the service (let's assume the service is the production of some music they'd like to listen to) pay a fixed cost in order to access the results of that service, and in return they are given the means to access that service (say a licensed copy of the music).
Of course, this system breaks down if freeloaders enter the system and start accessing the services without paying for them. You could rely on sufficient honest people in the system to appropriately pay the service provider enough money for them to eat (they'll probably see about 50p from each album sale so that requires approximately 20,000 album sales a year to get the same salary as could be achieved by working as a cleaner in McDonalds), but most people are only honest when the rules are clearly defined. So we need a clear rule about when it is appropriate to contribute to the service provider - in this case that would be the mechanical copyright law.
So, copyright law is an extremely fair way as stating that if someone wants to enjoy the fruits of someone's labours, they need to pay that person. It breaks down where large corporation acquire copyright and then overstep the mark in terms of collecting on their copyright (e.g. demanding money because someone possessed the means to copy copyrighted material, but where there is no evidence that they did so), but where someone has a reasonable expectation of payment for their services, it is difficult to come up with a fairer mechanism for ensuring that payment.
Uh no. "Image sensors", like eyes (incidentally, how many other image sensors do _you_ have) do an awfully large amount of preprocessing before the signal is sent to the brain. The signal is only similar to other nerve signals in its transmission media and encoding principles. The data itself is quite different (c.f. data transmission and voice transmission over a CDMA network).
You're right however, when you say that this still needs to be translated by the brain. You're assumption appears to be, however, that if you can't see at birth then you'll never be able to see. In truth, no more than approx 30% of congenital cases of blindness are cortical in origin, the vast majority being related to developmental problems in one or both eyes (microphthalmos, anophthalmos etc.) In cases involving cortical disfunction, then clearly this technique is unlikely to work, but where there is no cortical abnormality the brain remains plastic enough to learn to see no matter how late the signals start arriving.
One landmark case involved a patient who had cataracts from birth being given surgery to remove these cataracts in their 30s. The patient was initially able to see only colours and lights. After some time, the patient began to recognise shapes and eventually was able to process visual stimuli to a similar degree to a non-handicapped individual. Unfortunately, the patient did not adapt well to being able to see, finding it distracting and unpleasant. They also felt exluded from their community of visually impaired friends. The issue was not that the brain was unable to adapt functionally to visual stimuli, but the patient was behaviourally adapted to life without vision.
The upshot of this is that restoration of vision to the congenitally blind is not new and is limited by the brains adaptability. What is new is the range of conditions that could be treated by an advance such as this (for the non-classically trained amongst you, anopthalamos means being born without any eyes - try treating that with drops).
As much risk as in accidentally creating an evil super robot by throwing a load of old motherboards in a washing machine and putting it on spin cycle.
Don't listen to Michael Crichton (or Prince Charles for that matter): there is no grey goo.
I find if you mock out everything the unit tests run great. 100% pass rate every time. The coverage is a little low though.
Mocking out database access is fine when all you care about is using the resulting data (e.g. when testing post processing of the data). However, when testing the actual database interaction (e.g. SQL generation, SQL driven data processing, data access layer etc.) there is no substitute for backing on to a real database to uncover real errors.
Then all you need to worry about is cross database support, cross platform support, pre-population of the database, roll-back, comparison against expected results...
...wasn't it going to save the net or something ;)
Write a rebuttal or link to support your
recurse as necessary