I probably shouldn't respond, since your trollish post already got modded to -1, but there is one element you bring up that I probably should have addressed up front.
Sure Google wouldn't cooperate with the US government when ordered.
Obviously Google will comply with the law. And the vast majority of legal data requests from the government are good things: subpoenas and search warrants issued by competent courts acting in good faith to address real problems. But there's no way Google is going to provide unlimited access, and if the government were to try to order it (note that there is no current legal mechanism that would allow such an order; it would have to be new legislation), you can be certain Google would fight it tooth and nail.
They need three companies to activate that feature. The first one is Google, the second one is Facebook. We know the third one won't be Apple
Google would refuse, so if three are needed (why is that?), and assuming that Facebook would play ball, you need two more.
How do I know Google would refuse? I work for Google and anything like that would be so severely opposed by the culture at Google that there's just no way it would happen, even if management wanted it to -- and management wouldn't. Sergey Brin, in particular, would be up in arms, as would most of the senior technical staff and lots of the rest. Larry Page would also be opposed, but I don't think he'd throw the screaming fit I'd expect from Brin. About the only way it could happen is if it were forced by legislation, and it wouldn't happen quietly, the lobbying would be loud and ferocious. If it still somehow happened there would be a hundred Google Snowdens. Or a thousand. I'd be one of them (though I think I could do it without being caught or having to flee).
Speaking of Snowden, that's a great example. I was working for Google in 2013 when Snowden's leaks came out and the immediate reaction to the PRISM stuff was utter disbelief with a strong leavening of readiness to grab pitchforks if it were somehow remotely true. There were some really heated TGIFs (weekly company-wide meeting). Then we found that the the NSA was tapping fiber between data centers, and people calmed down since it meant Google wasn't cooperating... and immediately set about making sure that every bit of data flowing across Google networks was encrypted. We already had a great key management infrastructure in place and the "encrypt everything" project had been in progress for some time.
And when I say "immediately" I mean "faster than was realistically possible". Deadlines for full compliance were short and completely immovable. One of the teams I work with made heavy use of sharded MySQL (which unlike Bigtable provides transactional consistency) via JDBC, but the standard MySQL JDBC stack provides no mechanism for encryption and it wasn't feasible to just run it in a TLS tunnel. So the team had less than 30 days to design, build, test and deploy a secure replacement that integrated with Google's key management infrastructure. And note that it had to work at Google scale; thousands, if not tens of thousands, of queries per second. They did, at least, already have a secure substrate to use. Google's key management and secure networking infrastructure is great.
Close to the deadline, it was discovered that there was a nasty and very hard to debug race condition that caused intermittent deadlocks (IIRC; it was something like that). In desperation the team said that if they didn't get more time they might have to just shut down for a week or two. Since they built/ran the billing systems, which collect and distribute all the money and a shutdown would inevitably create losses in the tens if not hundreds of millions of dollars, they figured that would get them a postponement. The answer from management was that they might have to shut down for a week or two, the deadline was not moving. As it turned out some amazing heroics plus a fair amount of bubble gum and baling twine kept things going until they solved all the problems.
So... that's how Googlers feel about sharing information with the government. And if that really surprises you, then you don't know nerds.
People assume that since Google tracks a great deal of user information to use in targeted advertising that Googlers must not care much about privacy, but nothing could be further from the truth. Google tracks user data, but is extraordinarily careful to ensure that it doesn't leak, not even internally, and isn't used for other purposes. And it is not sold; to government or anyone else.
It's no accident that Google is not among the many, many companies who've suffered leakage/loss of user data (with the exception of whatever
Those who automate, will be perpetuating their own demise. Automate the jobs away... nobody will have a jobs... nobody has money... nobody can buy your goods/services... your automated whatever shuts down.
Nonsense. We just need to restructure the economy, which we can and will do, and we'll all be better off for it as the marginal cost of production falls to nearly zero. We just don't want it to happen too fast, because restructuring is hard.
I think it may have had to do with the fact that most of these required Spanish classes taught the Castellano dialect and not the Mexican one. Talk about useless.
Bah. The dialects aren't that different. Some minor differences in pronunciation (c & z, mostly). A few bits of vocabulary. But it's roughly equivalent to the difference between American and British English. Speakers of the mainstream forms of either dialect can communicate without more than minor difficulties with speakers of the other. There are more extreme sub-dialects which are more different, e.g. imagine a Scot with a strong brogue trying to talk to an Alabama redneck. Actually, I don't think the Spanish-speaking world has any differences quite that large. Well, maybe if you took a Spaniard who speaks a dialect that is half Basque (which is not only not related to Spanish; it's not latinate or related to any other known language) and put them with an Argentine gaucho.
I speak Mexican Spanish... specifically Jarocho Spanish (the sub-dialect of Veracruz and surrounding area; which has a lot in common with Cubano and other Caribbean variants), and I have absolutely no trouble getting around in Spain.
The point is to expose kids to something new that *some* of them will use again.
I think that's the less important part of the point. The more important is to expose all of them to the concepts and methods used in these various disciplines, to give them the tools of thought, and to give them a little understanding of the scope of human knowledge.
I remember in middle school Spanish was a requirement to graduate and go to high school.
That seems unlikely. It's more likely that a year of some foreign language was required. Perhaps your school only offered Spanish.
Yet most people that took that class left without knowing anything they didn't already know.
If that's true, it's because your teacher was terrible, but I suspect that it's nonsense. Oh, I'm sure that kids didn't finish a year of middle school Spanish being able to speak the language, but that is far from the same this as not having learned anything. Learning even the rudiments of a foreign language exposes you to a lot of important ideas. It teaches a little bit of cultural tolerance, understanding and respect. It actually teaches you a lot about English, since learning another grammar helps you to understand much better what parts of speech and sentence structure are.
I fully support the teaching of CS in schools for much the same reason. Familiarization with this central aspect of modern civilization is important. It doesn't matter that 95% of the kids will never go on to write a line of code professionally, it exposes all of them to the rudiments of computational logic, to "thinking" the way computers "think". As Joseph Campbell put it "Computers are like Old Testament gods; lots of rules and no mercy." It's useful for everyone to understand rule and process-oriented thinking, because they're all going to be interacting with computers for their entire lives[1].
Actually nearly all basic education is like this. It's about giving kids the tools of thinking, more than about teaching specific concepts. We teach everyone algebra not because everyone is going to spend their lives solving quadratic equations but because the mode of abstract thinking that is mathematics is really valuable. Lots of math, starting with algebra, is comprised of thinking of the form "Suppose there is a solution. What must that solution look like? Can we identify a solution that satisfies the requirements". It's being able to reason abstractly about unknowns. Abstract thinking is not natural to most people, which is exactly why math is so hard for them... but it's also exactly why it's so valuable that they learn it even if they never again use the exact thing that they learned.
What's sad is that so few educated people actually understand what education is about or why it matters (aside: I'm sure some cynic will chime in with something about how it's really to beat the creativity and curiosity out of kids and make them into obedient robots. I'm not going to deny that the approach to education that we use has a large dose of that, but that's distinct from the reasons the elements of the curriculum are important.)
[1] It's possible that AI research will progress to the point where computers think more like humans do. If that happens, it may undermine some of the value of teaching procedural programming. But I suspect that even when AI efforts achieve true intelligence it will be different in important ways from human intelligence, and we may well also continue using lots of non-AI software for its predictability and comprehensibility.
Look again at society. If you want to see how disproportionately white men can be and are punished simply for being male and white, go to any court but especially family court.
Black men in criminal court would beg to differ. And not just a little.
just because you dont get offended (neither do i) does not in anyway let the racists off the hook simply because you are white and they arent. they are still racist
True, but if you believe that all racism is equal, you're wrong. Racism (or other discrimination) by the powerful against the weak is much worse than racism by the weak against the powerful. If a wealthy or powerful black man discriminates against a lower-class white man, that's worse than when a lower-class white man discriminates against a wealthy or powerful black man. Similarly, a powerful white discriminating against a less powerful black (by far the more common case) is worse than the reverse.
BUT, thanks to social context and history, people of apparently equal levels of power are not actually of equal levels of power when they are of different races. In general, in the United States, a white man is more powerful than a black man, because society treats him better, police are more likely to believe him, etc. There are exceptions, of course, contexts in which the roles are reversed, or where race truly isn't relevant, but in most contexts, in the US, in 2017, there is an extra power that derives merely from being white, or male. It's not as large as it used to be, but it does still exist, and probably will for some generations yet.
Yes, this means that the more privileged classes must be more careful to avoid discriminating against less privileged classes than the reverse. If as a white man that bothers you, have some integrity and sack up. it's more than offset by the benefits you get from being a member of the privileged class (even though various affirmative action programs try to balance the scales in some specific areas by tilting the other way, they don't succeed overall). If you disagree, because you don't see the benefits of your own privilege, that's because you're blind to them, not because they don't exist.
If you really want to get past all this, to get to where we can all just be color-blind... so do I. But anyone who believes we're already there is either blind, insane or a racist. The only way to get to that color-blind society is to trudge tediously through a few more generations of the privileged going out of their way to try to de-privilege themselves and the unprivileged complaining whenever they notice a difference in levels of privilege. Yes, there will be a lot of people abusing this along the way. It'll be messy. It'll be complicated. It will result in all sorts of unfairnessess in every direction. Sorry, it's an imperfect world.
Do your part to make it better, not to compound the unfairness or to hold back getting to a more color-blind state by acting as if we're already there. We're not, and the quickest way forward is to acknowledge it openly and deal with it head on.
The USA already enforces its laws on the RoW, so it's reasonable for any other country to do the same.
No. It's reasonable for other countries to also reject the US's attempts to enforce its laws on their soil. Or, as my mom taught me when I was little "Two wrongs don't make a right".
A ton of research has been done on minimum wage and the consensus is that is has negligible effect on employment.
That's a serious misstatement of the minimum wage research. There has been a lot of research, and it has found that small, cautious increases in minimum wages tend to have negligible negative effect on employment. This isn't because minimum wages have no negative effect in general, it's because governments tend to avoid raising it so far that it becomes damaging.
This study shows that Seattle has gone too far.
It's also worth pointing out that past studies may have less relevance in the future, as much larger swaths of the unskilled labor market become vulnerable to automation. If employers are already considering automating minimum wage jobs away, even a mild increase in labor costs may cause them to make the switch.
I'm not generally a fan of minimum wage laws anyway, but I think right now is a particularly bad time to raise them. I think we're already just a few years away from a massive automation-driven economic restructuring, and while the overall impact of automation is going to be tremendously positive, it's also going to force a very tough adjustment on a lot of people. The faster that change comes, the harder it will hit and the harder it will be to deal with. Government should avoid economic manipulation that will accelerate it.
OTOH, though it's too radical for implementation in the very near future, eliminating the minimum wage and other welfare programs in favor of a universal basic income would lower labor costs and thereby delay automation. It wouldn't prevent it, but it would slow it down, and it would also make the transition much less painful.
Wrong.
More growth = faster death = more rot = CO2 balance pretty much where it was....OR we would not now be at a 1 million year maximum
Wrong. More growth does equal more rot, but it also means more gigatonnage of biomass at any given time. If the normal (pick a normal) tonnage of biomass is X and the amount (at equilibrium) with higher CO2 is Y, then the carbon in (Y-X) is actually removed from the ecosystem. You'll still have more rotting, meaning more turnover, but we're talking about an equilibrium state, so for every gram that is returned by rotting, another gram is absorbed by growth.
Note that this means that plant life acts as a carbon buffer. As CO2 levels rise, plant biomass rises, absorbing some of the excess carbon. As CO2 levels fall, plant biomass decreases, releasing the stored carbon. (This is all predicated on the assumption that plant biomass does vary with CO2 levels; which I don't actually know for certain, but makes sense.)
BTW, I read your posts. You identify as a G employee when it suits your interests (listen to me cuz I work there!). Other times you pretend to play it even handed when you clearly have your finger on the scale.
You're wrong. I call it like I see it. About the only concession I make to my employment is that I limit my comments around legal issues. I've said very little on this one, for example. There's a lot I'd like to say, but one hypothetical is as far as I think I should go.
Quick fact check: Amazon doesn't own The Washington Post, Jeff Bezos -- in his personal capacity -- does.
Distinction without difference... The sentiment expressed is still perfectly plausible:
Bezos despises Trump and is likely to steer WaPo to criticize the President above and beyond what would be fair
Bezos is likely to have purchased WaPo with tax-considerations paying at least a partial role — the other big concern, no doubt, was to save money on lobbying, by flat-out owning the biggest loudspeaker in the capital.
Also, Trump sees no distinction between himself and the businesses that he owns (and, as president, no particular reason to distinguish between himself as president and the businesses that he owns, else he'd have divested or set up a blind trust, as previous presidents have), and he has no problem using one of his businesses to browbeat or reward someone involved in another, so he assumes Bezos is the same.
I don't know about WaPo, but generally newspapers have ownership arrangements structured to ensure that the editorial board is independent and can't be influenced by the owner(s). I'd expect that's true in this case as well, or that some other mechanism is in place to ensure editorial independence. It's important.
Society is going to have to come to terms with this somehow
What is going to happen completely depends on profits. That doesn't mean we stop thinking logically about the fundamental untenability of it.
I don't see how whether the algorithms are used by corporations to generate profits or by any other entity for any other purpose affects what I'm talking about. I'm talking about the inherent opacity of ML-based algorithmic decisionmaking. The origin or purpose of the algorithms doesn't affect that. It would be the same if we were talking about a non-profit search engine (or anything else).
Yes, but this learning is captured by corporations. The input is overwhelmingly that of the people using it, little realizing the power of their contribution, getting very little accountability from the hoarder of their information.
in this case it's corporations. It could be non-profits, it could be governments (and *will* be governments). It could even be individuals. Access to data is needed to create training sets, but corporations are far from the only entities with access to large data sets.
Please stop commenting on your employers stories pretending to not have a dog in the hunt. It's a bad look dude.
I have no "dog in the hunt". Yes, my employer is Google (as I often point out, and as is mentioned prominently in my slashdot bio), but that really has no impact on my thought about the quality of the Hangouts app, or material design. If I thought it sucked, I'd say so.
This is essentially the tyranny of the closed algorithm.
It's worth pointing out that opening the algorithm wouldn't help, either. Although the original PageRank algorithm is open (published in a patent), I'm sure that today's "algorithm" contains a large dose of neural networks, which are opaque even if you can see all of the details of how they work.
Society is going to have to come to terms with this somehow. Human decisionmaking is just as opaque, in reality, but humans continually invent self-explanations for their decisions, and when we scale human decisionmaking processes we have to codify simple rules to provide some semblance of consistency across the many people doing the deciding. Those rules can be examined, and individual decisions can be checked against the rules. But human decisionmaking is wholly inadequate in terms of both scalability and speed to rank search engine results. Algorithmic decisionmaking is the only option. And when manually-designed algorithms fail to produce the desired results, machine learning is used to get them... and the result is algorithms that work measurably better in all the test cases, and in production use, but are almost as opaque as the decisionmaking of any individual human.
Yeah, the UI of Hangout "works"... but I hate it anyway:D Can not use it right now, as I would need to log of with my "youtube account" and log on with my "gmail/hangouts account" to find some specifics.
Oh, I thought we were talking about the mobile apps.
His point is that plant growth can absorb some of it, but not all or even most of it. Really, they can barely store any of it for very long.
No, that's not my primary point, and I actually disagree with the last part. My point is that plants *can* store some of it, in the form of increased biomass, and they can do it as long as CO2 levels stay high enough to maintain the "greener" equilibrium.
A subsidiary point is that this mechanism cannot absorb all of the carbon. Assuming that living plant biomass is correlated directly with CO2 level in the atmosphere (because CO2 promotes photosynthesis and therefore growth), absorbing all the excess CO2 would reduce CO2 level, and therefore reduce living plant biomass. The now-dead excess plant biomass would release most of its carbon back into the atmosphere, thereby raising the CO2 level. So wherever the new equilibrium point is, it will be at higher than "normal" CO2 levels.
Yeah I think it is a misunderstanding of biology that forests and grasslands are a carbon sink. Forests just rot and release carbon in a state of equilibrium.
In a state of equilibrium, yes. If high CO2 levels cause vegetation to get more extensive, thicker, etc., it creates a new equilibrium with more gigatons of carbon tied up in living plants, thus removing those gigatons of carbon from the atmosphere. This obviously cannot lower atmospheric carbon concentrations back to pre-industrial levels, because assuming it can drop CO2 levels signficantly, as soon as they drop to the point that the vegetation level starts to die back, the carbon tied up in the dead plants gets released back in to the atmosphere (most of it, anyway).
I'm not suggesting that we can just sit back and assume that plant life will solve our problem for us. It won't. But it likely will soak up some of the excess carbon.
Not a huge fan of turning our oceans and atmosphere into a 100% man-managed ecosystem
I'm a huge fan of exactly that. Further, I think that it is necessary in the long term. I think we're very, very far from knowing how to do it, but that we should start learning. Now. We've already unintentionally altered the planetary environment, now we should learn to do it on purpose and in the way we want.
[Haida is] a promising result.
I don't know about that. It wasn't a well-conducted scientific experiment. We can, and should, do much better. Still, the basic concept is good.
Slashdot Mirror
I probably shouldn't respond, since your trollish post already got modded to -1, but there is one element you bring up that I probably should have addressed up front.
Sure Google wouldn't cooperate with the US government when ordered.
Obviously Google will comply with the law. And the vast majority of legal data requests from the government are good things: subpoenas and search warrants issued by competent courts acting in good faith to address real problems. But there's no way Google is going to provide unlimited access, and if the government were to try to order it (note that there is no current legal mechanism that would allow such an order; it would have to be new legislation), you can be certain Google would fight it tooth and nail.
They need three companies to activate that feature. The first one is Google, the second one is Facebook. We know the third one won't be Apple
Google would refuse, so if three are needed (why is that?), and assuming that Facebook would play ball, you need two more.
How do I know Google would refuse? I work for Google and anything like that would be so severely opposed by the culture at Google that there's just no way it would happen, even if management wanted it to -- and management wouldn't. Sergey Brin, in particular, would be up in arms, as would most of the senior technical staff and lots of the rest. Larry Page would also be opposed, but I don't think he'd throw the screaming fit I'd expect from Brin. About the only way it could happen is if it were forced by legislation, and it wouldn't happen quietly, the lobbying would be loud and ferocious. If it still somehow happened there would be a hundred Google Snowdens. Or a thousand. I'd be one of them (though I think I could do it without being caught or having to flee).
Speaking of Snowden, that's a great example. I was working for Google in 2013 when Snowden's leaks came out and the immediate reaction to the PRISM stuff was utter disbelief with a strong leavening of readiness to grab pitchforks if it were somehow remotely true. There were some really heated TGIFs (weekly company-wide meeting). Then we found that the the NSA was tapping fiber between data centers, and people calmed down since it meant Google wasn't cooperating... and immediately set about making sure that every bit of data flowing across Google networks was encrypted. We already had a great key management infrastructure in place and the "encrypt everything" project had been in progress for some time.
And when I say "immediately" I mean "faster than was realistically possible". Deadlines for full compliance were short and completely immovable. One of the teams I work with made heavy use of sharded MySQL (which unlike Bigtable provides transactional consistency) via JDBC, but the standard MySQL JDBC stack provides no mechanism for encryption and it wasn't feasible to just run it in a TLS tunnel. So the team had less than 30 days to design, build, test and deploy a secure replacement that integrated with Google's key management infrastructure. And note that it had to work at Google scale; thousands, if not tens of thousands, of queries per second. They did, at least, already have a secure substrate to use. Google's key management and secure networking infrastructure is great.
Close to the deadline, it was discovered that there was a nasty and very hard to debug race condition that caused intermittent deadlocks (IIRC; it was something like that). In desperation the team said that if they didn't get more time they might have to just shut down for a week or two. Since they built/ran the billing systems, which collect and distribute all the money and a shutdown would inevitably create losses in the tens if not hundreds of millions of dollars, they figured that would get them a postponement. The answer from management was that they might have to shut down for a week or two, the deadline was not moving. As it turned out some amazing heroics plus a fair amount of bubble gum and baling twine kept things going until they solved all the problems.
So... that's how Googlers feel about sharing information with the government. And if that really surprises you, then you don't know nerds.
People assume that since Google tracks a great deal of user information to use in targeted advertising that Googlers must not care much about privacy, but nothing could be further from the truth. Google tracks user data, but is extraordinarily careful to ensure that it doesn't leak, not even internally, and isn't used for other purposes. And it is not sold; to government or anyone else.
It's no accident that Google is not among the many, many companies who've suffered leakage/loss of user data (with the exception of whatever
Those who automate, will be perpetuating their own demise. Automate the jobs away... nobody will have a jobs... nobody has money... nobody can buy your goods/services... your automated whatever shuts down.
Nonsense. We just need to restructure the economy, which we can and will do, and we'll all be better off for it as the marginal cost of production falls to nearly zero. We just don't want it to happen too fast, because restructuring is hard.
I think it may have had to do with the fact that most of these required Spanish classes taught the Castellano dialect and not the Mexican one. Talk about useless.
Bah. The dialects aren't that different. Some minor differences in pronunciation (c & z, mostly). A few bits of vocabulary. But it's roughly equivalent to the difference between American and British English. Speakers of the mainstream forms of either dialect can communicate without more than minor difficulties with speakers of the other. There are more extreme sub-dialects which are more different, e.g. imagine a Scot with a strong brogue trying to talk to an Alabama redneck. Actually, I don't think the Spanish-speaking world has any differences quite that large. Well, maybe if you took a Spaniard who speaks a dialect that is half Basque (which is not only not related to Spanish; it's not latinate or related to any other known language) and put them with an Argentine gaucho.
I speak Mexican Spanish... specifically Jarocho Spanish (the sub-dialect of Veracruz and surrounding area; which has a lot in common with Cubano and other Caribbean variants), and I have absolutely no trouble getting around in Spain.
The point is to expose kids to something new that *some* of them will use again.
I think that's the less important part of the point. The more important is to expose all of them to the concepts and methods used in these various disciplines, to give them the tools of thought, and to give them a little understanding of the scope of human knowledge.
I remember in middle school Spanish was a requirement to graduate and go to high school.
That seems unlikely. It's more likely that a year of some foreign language was required. Perhaps your school only offered Spanish.
Yet most people that took that class left without knowing anything they didn't already know.
If that's true, it's because your teacher was terrible, but I suspect that it's nonsense. Oh, I'm sure that kids didn't finish a year of middle school Spanish being able to speak the language, but that is far from the same this as not having learned anything. Learning even the rudiments of a foreign language exposes you to a lot of important ideas. It teaches a little bit of cultural tolerance, understanding and respect. It actually teaches you a lot about English, since learning another grammar helps you to understand much better what parts of speech and sentence structure are.
I fully support the teaching of CS in schools for much the same reason. Familiarization with this central aspect of modern civilization is important. It doesn't matter that 95% of the kids will never go on to write a line of code professionally, it exposes all of them to the rudiments of computational logic, to "thinking" the way computers "think". As Joseph Campbell put it "Computers are like Old Testament gods; lots of rules and no mercy." It's useful for everyone to understand rule and process-oriented thinking, because they're all going to be interacting with computers for their entire lives[1].
Actually nearly all basic education is like this. It's about giving kids the tools of thinking, more than about teaching specific concepts. We teach everyone algebra not because everyone is going to spend their lives solving quadratic equations but because the mode of abstract thinking that is mathematics is really valuable. Lots of math, starting with algebra, is comprised of thinking of the form "Suppose there is a solution. What must that solution look like? Can we identify a solution that satisfies the requirements". It's being able to reason abstractly about unknowns. Abstract thinking is not natural to most people, which is exactly why math is so hard for them... but it's also exactly why it's so valuable that they learn it even if they never again use the exact thing that they learned.
What's sad is that so few educated people actually understand what education is about or why it matters (aside: I'm sure some cynic will chime in with something about how it's really to beat the creativity and curiosity out of kids and make them into obedient robots. I'm not going to deny that the approach to education that we use has a large dose of that, but that's distinct from the reasons the elements of the curriculum are important.)
[1] It's possible that AI research will progress to the point where computers think more like humans do. If that happens, it may undermine some of the value of teaching procedural programming. But I suspect that even when AI efforts achieve true intelligence it will be different in important ways from human intelligence, and we may well also continue using lots of non-AI software for its predictability and comprehensibility.
This is just a scam and play on words to redefine racism as something which can only done by white people.
You didn't read the post you replied to.
You could go on forever with that logic, but it's ultimately self-defeating.
Exactly wrong. Correcting the imbalances is the only way to overcome them.
You're wrong. As long as we allow existing imbalances to be perpetuated, we'll never break free.
Look again at society. If you want to see how disproportionately white men can be and are punished simply for being male and white, go to any court but especially family court.
Black men in criminal court would beg to differ. And not just a little.
just because you dont get offended (neither do i) does not in anyway let the racists off the hook simply because you are white and they arent. they are still racist
True, but if you believe that all racism is equal, you're wrong. Racism (or other discrimination) by the powerful against the weak is much worse than racism by the weak against the powerful. If a wealthy or powerful black man discriminates against a lower-class white man, that's worse than when a lower-class white man discriminates against a wealthy or powerful black man. Similarly, a powerful white discriminating against a less powerful black (by far the more common case) is worse than the reverse.
BUT, thanks to social context and history, people of apparently equal levels of power are not actually of equal levels of power when they are of different races. In general, in the United States, a white man is more powerful than a black man, because society treats him better, police are more likely to believe him, etc. There are exceptions, of course, contexts in which the roles are reversed, or where race truly isn't relevant, but in most contexts, in the US, in 2017, there is an extra power that derives merely from being white, or male. It's not as large as it used to be, but it does still exist, and probably will for some generations yet.
Yes, this means that the more privileged classes must be more careful to avoid discriminating against less privileged classes than the reverse. If as a white man that bothers you, have some integrity and sack up. it's more than offset by the benefits you get from being a member of the privileged class (even though various affirmative action programs try to balance the scales in some specific areas by tilting the other way, they don't succeed overall). If you disagree, because you don't see the benefits of your own privilege, that's because you're blind to them, not because they don't exist.
If you really want to get past all this, to get to where we can all just be color-blind... so do I. But anyone who believes we're already there is either blind, insane or a racist. The only way to get to that color-blind society is to trudge tediously through a few more generations of the privileged going out of their way to try to de-privilege themselves and the unprivileged complaining whenever they notice a difference in levels of privilege. Yes, there will be a lot of people abusing this along the way. It'll be messy. It'll be complicated. It will result in all sorts of unfairnessess in every direction. Sorry, it's an imperfect world.
Do your part to make it better, not to compound the unfairness or to hold back getting to a more color-blind state by acting as if we're already there. We're not, and the quickest way forward is to acknowledge it openly and deal with it head on.
The USA already enforces its laws on the RoW, so it's reasonable for any other country to do the same.
No. It's reasonable for other countries to also reject the US's attempts to enforce its laws on their soil. Or, as my mom taught me when I was little "Two wrongs don't make a right".
A ton of research has been done on minimum wage and the consensus is that is has negligible effect on employment.
That's a serious misstatement of the minimum wage research. There has been a lot of research, and it has found that small, cautious increases in minimum wages tend to have negligible negative effect on employment. This isn't because minimum wages have no negative effect in general, it's because governments tend to avoid raising it so far that it becomes damaging.
This study shows that Seattle has gone too far.
It's also worth pointing out that past studies may have less relevance in the future, as much larger swaths of the unskilled labor market become vulnerable to automation. If employers are already considering automating minimum wage jobs away, even a mild increase in labor costs may cause them to make the switch.
I'm not generally a fan of minimum wage laws anyway, but I think right now is a particularly bad time to raise them. I think we're already just a few years away from a massive automation-driven economic restructuring, and while the overall impact of automation is going to be tremendously positive, it's also going to force a very tough adjustment on a lot of people. The faster that change comes, the harder it will hit and the harder it will be to deal with. Government should avoid economic manipulation that will accelerate it.
OTOH, though it's too radical for implementation in the very near future, eliminating the minimum wage and other welfare programs in favor of a universal basic income would lower labor costs and thereby delay automation. It wouldn't prevent it, but it would slow it down, and it would also make the transition much less painful.
Wrong. More growth = faster death = more rot = CO2 balance pretty much where it was....OR we would not now be at a 1 million year maximum
Wrong. More growth does equal more rot, but it also means more gigatonnage of biomass at any given time. If the normal (pick a normal) tonnage of biomass is X and the amount (at equilibrium) with higher CO2 is Y, then the carbon in (Y-X) is actually removed from the ecosystem. You'll still have more rotting, meaning more turnover, but we're talking about an equilibrium state, so for every gram that is returned by rotting, another gram is absorbed by growth.
Note that this means that plant life acts as a carbon buffer. As CO2 levels rise, plant biomass rises, absorbing some of the excess carbon. As CO2 levels fall, plant biomass decreases, releasing the stored carbon. (This is all predicated on the assumption that plant biomass does vary with CO2 levels; which I don't actually know for certain, but makes sense.)
BTW, I read your posts. You identify as a G employee when it suits your interests (listen to me cuz I work there!). Other times you pretend to play it even handed when you clearly have your finger on the scale.
You're wrong. I call it like I see it. About the only concession I make to my employment is that I limit my comments around legal issues. I've said very little on this one, for example. There's a lot I'd like to say, but one hypothetical is as far as I think I should go.
Distinction without difference... The sentiment expressed is still perfectly plausible:
Also, Trump sees no distinction between himself and the businesses that he owns (and, as president, no particular reason to distinguish between himself as president and the businesses that he owns, else he'd have divested or set up a blind trust, as previous presidents have), and he has no problem using one of his businesses to browbeat or reward someone involved in another, so he assumes Bezos is the same.
I don't know about WaPo, but generally newspapers have ownership arrangements structured to ensure that the editorial board is independent and can't be influenced by the owner(s). I'd expect that's true in this case as well, or that some other mechanism is in place to ensure editorial independence. It's important.
Society is going to have to come to terms with this somehow
What is going to happen completely depends on profits. That doesn't mean we stop thinking logically about the fundamental untenability of it.
I don't see how whether the algorithms are used by corporations to generate profits or by any other entity for any other purpose affects what I'm talking about. I'm talking about the inherent opacity of ML-based algorithmic decisionmaking. The origin or purpose of the algorithms doesn't affect that. It would be the same if we were talking about a non-profit search engine (or anything else).
Yes, but this learning is captured by corporations. The input is overwhelmingly that of the people using it, little realizing the power of their contribution, getting very little accountability from the hoarder of their information.
in this case it's corporations. It could be non-profits, it could be governments (and *will* be governments). It could even be individuals. Access to data is needed to create training sets, but corporations are far from the only entities with access to large data sets.
You're free to remind everyone that I work for Google. I often point it out myself, when I think it's relevant.
FWIW, my job isn't fixing Android security bugs. Other people do that. My job is building new crypto tools for app developers to use.
Please stop commenting on your employers stories pretending to not have a dog in the hunt. It's a bad look dude.
I have no "dog in the hunt". Yes, my employer is Google (as I often point out, and as is mentioned prominently in my slashdot bio), but that really has no impact on my thought about the quality of the Hangouts app, or material design. If I thought it sucked, I'd say so.
This is essentially the tyranny of the closed algorithm.
It's worth pointing out that opening the algorithm wouldn't help, either. Although the original PageRank algorithm is open (published in a patent), I'm sure that today's "algorithm" contains a large dose of neural networks, which are opaque even if you can see all of the details of how they work.
Society is going to have to come to terms with this somehow. Human decisionmaking is just as opaque, in reality, but humans continually invent self-explanations for their decisions, and when we scale human decisionmaking processes we have to codify simple rules to provide some semblance of consistency across the many people doing the deciding. Those rules can be examined, and individual decisions can be checked against the rules. But human decisionmaking is wholly inadequate in terms of both scalability and speed to rank search engine results. Algorithmic decisionmaking is the only option. And when manually-designed algorithms fail to produce the desired results, machine learning is used to get them... and the result is algorithms that work measurably better in all the test cases, and in production use, but are almost as opaque as the decisionmaking of any individual human.
Yeah, the UI of Hangout "works" ... but I hate it anyway :D Can not use it right now, as I would need to log of with my "youtube account" and log on with my "gmail/hangouts account" to find some specifics.
Oh, I thought we were talking about the mobile apps.
I'm not sure why you say "except". I never denied that it was already happening. I didn't know that it was, but it makes sense.
His point is that plant growth can absorb some of it, but not all or even most of it. Really, they can barely store any of it for very long.
No, that's not my primary point, and I actually disagree with the last part. My point is that plants *can* store some of it, in the form of increased biomass, and they can do it as long as CO2 levels stay high enough to maintain the "greener" equilibrium.
A subsidiary point is that this mechanism cannot absorb all of the carbon. Assuming that living plant biomass is correlated directly with CO2 level in the atmosphere (because CO2 promotes photosynthesis and therefore growth), absorbing all the excess CO2 would reduce CO2 level, and therefore reduce living plant biomass. The now-dead excess plant biomass would release most of its carbon back into the atmosphere, thereby raising the CO2 level. So wherever the new equilibrium point is, it will be at higher than "normal" CO2 levels.
Yeah I think it is a misunderstanding of biology that forests and grasslands are a carbon sink. Forests just rot and release carbon in a state of equilibrium.
In a state of equilibrium, yes. If high CO2 levels cause vegetation to get more extensive, thicker, etc., it creates a new equilibrium with more gigatons of carbon tied up in living plants, thus removing those gigatons of carbon from the atmosphere. This obviously cannot lower atmospheric carbon concentrations back to pre-industrial levels, because assuming it can drop CO2 levels signficantly, as soon as they drop to the point that the vegetation level starts to die back, the carbon tied up in the dead plants gets released back in to the atmosphere (most of it, anyway).
I'm not suggesting that we can just sit back and assume that plant life will solve our problem for us. It won't. But it likely will soak up some of the excess carbon.
Not a huge fan of turning our oceans and atmosphere into a 100% man-managed ecosystem
I'm a huge fan of exactly that. Further, I think that it is necessary in the long term. I think we're very, very far from knowing how to do it, but that we should start learning. Now. We've already unintentionally altered the planetary environment, now we should learn to do it on purpose and in the way we want.
[Haida is] a promising result.
I don't know about that. It wasn't a well-conducted scientific experiment. We can, and should, do much better. Still, the basic concept is good.