I don't think so. If I understand the law correctly, there's no whistleblower exemption for what he did, so any attempt to make that sort of an argument to the jury will be met with a big fat, "Shut the fuck up" from the judge. It's like medical marijuana sellers trying to defend themselves in federal court. They're not allowed to even say that they were selling it for medical purposes because federal law makes no distinction. The only question is whether they were selling weed, not why.
I strongly suspect that if the penalty for not going to the back of the bus was life in prison or execution, Rosa Parks might have decided that the back of the bus was good enough.
There's a simple reason for that: The price of a Big Mac is probably driven mostly by volatile inputs like food and energy. Food an energy are only a part of the overall inflation picture, and they're generally not leading indicators of long term trends. There are two key CPI numbers: headline and core. Core removes volatile components like food and energy because they tend not to be built into long-term inflation expectations. When headline and core diverge significantly, the divergence generally goes away in the near future, and it generally goes away when the headline number moves back toward the core number, not the other way around.
Long term, inflation is driven mostly by expectations and the interaction between expectations and wages. Spikes or drops in the price of oil or beef tend to revert to historical norms, so while they make for interesting charts in the news, they're not really all that useful for long-run predictions because they don't really provide steady enough "feedback" to feed into slower moving prices like wages. The divergence between the Big Mac index and our other inflation metrics is likely driven by that phenomenon rather than an actual failure of the broader metrics.
Anyway, it's not a matter of "assuming" the BLS knows best. You'll find that people who actually study this stuff and use the data think they put out a very useful set of indices and have very good reasons to ignore outliers like the Big Mac index. The BLS basket of goods is very broad, well analyzed and completely public. Every quarter we hear the big headline about something like, "Chicken prices spiraling out of control! Inflation to come!" Not unless the public at large *really* eats tons and tons of chicken and the trend continues for some time.
There are other cross-checks that are pretty easy to do. For example, if the Big Mac Index was truly reflective of reality, we'd be seeing massive capital flight from the US to foreign markets with better inflation numbers. If you know the dollar is losing 10% in inflation every year, just sell your dollar-denominated assets, buy Japanese bonds with yen and enjoy your nearly risk free ~12% real return in dollars. Either the big money (who are presumably the puppet masters driving this whole scam) is too dumb to do this or that's really not how the numbers work out. Either that or all currencies everywhere are inflating at roughly the same rate without feeding back into wages anywhere, but that would be a really interesting macroeconomic state of affairs.
The intersection of "smart enough to be careful when planning your assassination" and "dumb / crazy enough to try assassinating a sitting US President" is probably very, very small. No doubt it's dwarfed by several orders of magnitude by the set "dumb enough to threaten to kill the President."
Comapring data points between two workers has all sort sof problems as you point out. But having all of the salary information public for everybody would lead to a much better estimate of the market price overall, which would be a good thing. Employers make arguments based on their individual quantization error and statistical outliers, but the real benefit is in the aggregate information. Knowing not just what your neighbor makes but whate everybody everywhere (including competing companies) makes allows you to make much better career decisions, the individual details about the guy in the next cube notwithstanding.
If we made that change, there would be a massive reshuffiling right off the bat, probably followed by a return to stability. If you have an individual case where you needed to make a weird lowball offer to somebody, they'd have the information needed to decide whether to take it up front. If you needed to pay more for a skill that was in demand on a short timeline, that would probably reflected in total market prices. Instead of having to pay your current people more to match the new employee's rate, you'd probably already have lost your employees who were making less than the market rate. Which is probably a good thing as well, because there's clearly plenty of high value work for them to be doing.
Take a look at the BLS data instead of sampling single data points. Like most people, your estimate of inflation is based on your gut and not on real data. Think about it this way: Assuming you're spot-on about soda and the price is climbing rapidly, what percentage of the average American household income goes to buying soda? How would even a small change in the price of some other good affect the usefulness of that data point?
But by the iPad Gen 1 index, inflation is way down. By the gold index, we're in deflation. By the California water index, inflation is out of control. But those aren't really indexes. They're just prices. The BLS puts a *ton* of effort into creating an index of what people actually buy that's comparable from year to year. A single price randomly sampled just won't do it. If you ask the average American, they seem to think they spend 95% of their income on gasoline and milk. But they don't. They spend a moderate percentage of their income on gas and hardly any of it on milk. They probably don't remember the 1/10th of a refrigerator they "bought" this year due to depreciation, for example.
The reality is that the BLS numbers match up well with the independent billion price index and with inflation expectations built into the financial markets. The odds that shadowstats or some other bogus index is the "real" index are slim to none. If we were really getting hammered at the the rates that, say, shadowstats, would have us believe, we'd all be destitute and we'd have noticed at least some captial flows following the money.
That's even better. Knowing what the guy in the next cube over makes just makes for a short-term rejiggering of the local salary distribution. Knowing with fairly high precision what you could be making literally anywhere else actually makes for good market rates. GlassDoor is a good start, but the market for employees would be much more efficient if all of that information was public and accurate. It would feel "icky" to us for a few months because of our culture and it would piss off corporate management (generally the benficiary of the current inefficiency), but the long run equilbiriubm would be better.
The grand problem isn't about a comparison between two employees and trying to justify why their salaries aren't identical. The real issue is that all of the salary information being public would actually result in good price setting information for everybody. Once we got over the initial shock that everybody doesn't make exactly the same amount of money, employees would actually be able to look at the aggregates and the data for other companies and make rational decisions about what they're really worth. As it stands, people are just guessing and a lot of suboptimal pricing is happening.
This is something that never made sense to me about the legal profession. If an engineer designed a bridge that violated building codes, he'd be in a world of shit. If a pilot ingored FAA regulations or safety best practices, that would be the end of his career. A lawyer can write a contract that's 80% unenforceable nonsense, sell that service to his employer as good legal work, and walk away with a paycheck and a blessing from his guild.
The non-union version would be a little different, though. Employers would have to pay their underpaid good performers more, and if budget started to get tight as salaries normalized, they'd have to get rid of the drooling idiot who wasn't earning his keep. In the union version, it's hard to get rid of the unproductive idiot for anything short of murder. In a competitive workplace, it's just a question of total salary budget.
That's a problem with management giving realistic assessments and comparisons to their employees, though. If you think you're above average and you're not, your manager should be able to give you a good reason why you're not. Ultimately after the initial shake up (even if it involves people moving to other companies where they'll *still* likely be rated as below average), people will settle into reality.
The thought experiments people do about this stuff seems to conflict with real experiments. The NBA doesn't collapse into chaos because the news reports it whenever LeBron James gets a new contract. Other players know that they're making less than he is and they probably generally know why. It's just part of the culture, and the equilbrium dosn't seem to be any differnt from a workplace where salaries are kept a closely guarded secret. It would just take time for our culture to shift.
It doesn't "have" to be perfect to be a working market. It just has to be perfect in order to be a completely idealized perfectly competitive free market on paper. No markets live up to the perfectly idealized theory, but they generally work. They just work less well as you move away from that ideal state. "I don't know what my competitor whill charge tomorrow" is a slight deviation from perfect information, but it's not really a big deal. "Nobody knows what anybody else is charging," is a major deviation that causes real problems.
It's a problem that applies really well to the medical market as well: You don't know nearly as well as your doctor what the actual value of his services are. You also often don't know what he'll charge you until after you've "bought" his services. And you sure as hell don't have a list of what other doctors would charge you. The end result is that the market is a total mess. Every step away from that problem and toward more perfect information (clearly published prices, more educated consumers, etc.) will improve outcomes, even if we never get to true "perfect information."
1) Different students might need to be taught in different ways to learn the same set of skills.
2) Different students should come out of school with very different sets of skills.
I'm totally onboard with (1), but not so much with (2). "He can't add, but boy can he write!" Everybody is going to have natural aptitudes for certain things, but there's a basic level of core competence that everybody should be able to hit.
Since non-GMO foods have been consumed over many lifetimes of consumption, their safety or lack thereof has been pretty well established.
We're constantly hybridizing and producing all manner of crazy new plants where God knows how many genes are combined in totally new ways. There are tons of hybrids we're eating right now that have nowhere near "many lifetimes" worth of consumption, and there's no good reason to think that we'll do a better job of predicting all of their subtle properties (especially "long term" safety) any better than if we take a known organism and add one carefully selected gene to it.
I'm in favor of good information on food labels, but I also see the complaint from the GMO producers. I think they're worried about something like this:
GMO company: There's no need to label these things. They're perfectly safe.
Anti-GMO activist: Why do you hate transparency? If it's perfectly safe, there's no reason not to label them.
[Time passes. Labels mandated.]
Anti-GMO activist: If GMOs are so safe, why is labeling them mandatory?
Consumers: Hey! That's a good point!
Same here. We think and solve problems similarly, so there isn't much "talking past each other" when we disagree. If a problem can be reasoned through, we usually come to agreement. If it's squishier, we both sort of recognize that it's probably not worth fighting to the death over. Money and worry about unemployment is never a problem, so we don't fight over it or nitpick how it's spent (in fact, we each have a mix of private and shared accounts after 10 years of marriage, and we each handle a subset of the household financees without much oversight from the other). Beyond that I suppose everything else is pretty normal, but those are two big issues that cause a lot of relationship trouble that work out to a pretty strong built-in advantage for us.
We're both pretty mellow people who don't usually get wound up about stuff that's not really important, but I'm not sure if that's just who we are or if it's associated with how we were educated and how we work. There seem to be certain types of people who genuinely get bored and need a certain amount of drama in their human interactions and will create it if it's not there. Those people seem to be less common in engineering, although I don't have a lot to quantify or support that. Those people are usually not fun to be around and seem like they'd be a nightmare to be married to.
I'm all for a massive reduction in police militarization and very nearly eliminating SWAT teams given how rarely they're used for what they were originally intended for. That being said, the problem with using this issue as a lever is that there will always be some fringe situation that calls for a swift armed reaction of some sort. These kids can literally say whatever they want, so there's always going to be a way to provoke a dangerous response as long as there's any conceivable situation that warrants that response. This is just one of those cases where there should be a panic button. It should be really difficult to hit and you should punish the living shit out of people who treat the panic button like it's a toy.
Let me step back a bit and rephrase my earlier question. There are two key price points: There's the upper limit, which is the amount of money you make for your employer. They wouldn't pay you a dollar more than that because they'd just be losing a dollar on the net, and they'd be indifferent between hiring you and not hiring you if it equals that amount because the net value to them is $0. Then there's the minimum amount you'd work for. You wouldn't work for less than that, by definition. So now there's a big range between those two numbers.
The salary you land on in that range reflects a split of that surplus value between you and your employer. You seem to think that you getting some of it while your employer keeps some of it is reasonable. You also seem to think that you getting none of it while your employer keeps all of it is reasonable. But somewhere along the line, you seem to think that you taking "too much" of it becomes unreasonable. I'm wondering what you're basing that on and how you decide when "too much" has happened. Why the asymmetry in favor of the employer, and why the certainty that Shearer crossed it somewhere between $300K and $450K per episode?
So if your employer cut you back to that base requirement, you wouldn't consider leaving and using your time some other way? Or at least work fewer hours?
Imagine you had enough money to live your current standard of living forever without working any more. Would that change your base requirement? You have a set of choices with respect to what to do with your time. One option would be not to work at all and to spend time with friends and family and enriching hobbies. Assuming that was an option to you, would the base pay requirement to get you to give that time up and go to work every day go up, or would it stay the same regardless of whether you needed the money or not?
I ask because Shearer has tons of money. A paycheck that would rent him a cheap apartment and a daily ration of Top Ramen would not affect his standard of living at all. He wouldn't notice either way. So given that he's probably a normal person with friends and family and hobbies, he'd be crazy to go to work at a job that only pays enough for a cheap apartment and a daily ration of Top Ramen. He's better off enjoying the days he has left doing something more enriching. Being "grateful" that an employer offered him a subsistence wage in exchange for losing that time wouldn't be rational at all.
If you had gotten paid a larger chunk of that money, would it have been morally wrong somehow?
Are you sure you deserved your whole salary as a software developer? I bet you'd have done it for less than you were paid, so why did you take that money? What are you really worth, and how do you know?
Perhaps you could explain how such a salary is justified? Without resorting to "well the Market says..."?
Becuse he's earning his employers shittons more than $6M per year by doing that work? Would the morally correct outcome be for him to cut his salary by, say, $3M so the owners of the company can pocket $3M more?
Monopoly or not, companies set their prices at the spot that they think maximizes their profits. They can't "pass on" 100% of a new cost by changing the price because that knocks them off of the spot they optimized for. They do lose something in the process. If all it took for them to grab an extra $229K in profits was to raise their prices, they'd have done it already.
One could point out that there are fewer instances of white males being miscategorized.
White males are just about the easiest faces to categorize. They tend to have short hair that doesn't obscure facial features or create oddball shapes that confuse the classifiers. Their skin tone makes photographing them and finding edges extracting features easier than it is with darker skinned people. White people have a greater variety of eye colors that can be used to distinguish among them. "White guy face" is just about the optimal case for this problem. If I had to come up with a worst case that was also a photo of a fairly "common" person, I'd go with "dark skinned, brown eyed person wtih long hair and facial hair." That's a pretty clean sweep of all of the variables that make this a hard problem.
Good decisions? Sorry, but releasing poorly tested software like this was obviously a bad decision. The bad outcomes were a direct result of their poor decision making.
How good does the cutting edge of object recognition need to be before it's not "poorly tested" anymore, especially when it's for a silly photo app and not a medical or military application? I never hear this type of thing from people who have actually had to solve these types of problems. The reality is that objects are going to be confused with other objects. Lots of them, once we're talking about hundreds of millions or billions of samples. Some cases will fail with great regularity and patterns. The unfortunate fact here was that the pattern happened to coincidentally have really embarrassing cultural connotations.
This is one of the things I don't miss about working in machine vision. We'd run our algorithm over a zillion images and it would correctly handle all of them save a small handful and that small handful would be filed as bugs. OK, maybe we'll be able to handle that small handful at the expense of a smaller handful next time around. But the pass/fail criteria for the tool is in its overall results, not in the outliers.
Slashdot Mirror
I don't think so. If I understand the law correctly, there's no whistleblower exemption for what he did, so any attempt to make that sort of an argument to the jury will be met with a big fat, "Shut the fuck up" from the judge. It's like medical marijuana sellers trying to defend themselves in federal court. They're not allowed to even say that they were selling it for medical purposes because federal law makes no distinction. The only question is whether they were selling weed, not why.
I strongly suspect that if the penalty for not going to the back of the bus was life in prison or execution, Rosa Parks might have decided that the back of the bus was good enough.
There's a simple reason for that: The price of a Big Mac is probably driven mostly by volatile inputs like food and energy. Food an energy are only a part of the overall inflation picture, and they're generally not leading indicators of long term trends. There are two key CPI numbers: headline and core. Core removes volatile components like food and energy because they tend not to be built into long-term inflation expectations. When headline and core diverge significantly, the divergence generally goes away in the near future, and it generally goes away when the headline number moves back toward the core number, not the other way around.
Long term, inflation is driven mostly by expectations and the interaction between expectations and wages. Spikes or drops in the price of oil or beef tend to revert to historical norms, so while they make for interesting charts in the news, they're not really all that useful for long-run predictions because they don't really provide steady enough "feedback" to feed into slower moving prices like wages. The divergence between the Big Mac index and our other inflation metrics is likely driven by that phenomenon rather than an actual failure of the broader metrics.
Anyway, it's not a matter of "assuming" the BLS knows best. You'll find that people who actually study this stuff and use the data think they put out a very useful set of indices and have very good reasons to ignore outliers like the Big Mac index. The BLS basket of goods is very broad, well analyzed and completely public. Every quarter we hear the big headline about something like, "Chicken prices spiraling out of control! Inflation to come!" Not unless the public at large *really* eats tons and tons of chicken and the trend continues for some time. There are other cross-checks that are pretty easy to do. For example, if the Big Mac Index was truly reflective of reality, we'd be seeing massive capital flight from the US to foreign markets with better inflation numbers. If you know the dollar is losing 10% in inflation every year, just sell your dollar-denominated assets, buy Japanese bonds with yen and enjoy your nearly risk free ~12% real return in dollars. Either the big money (who are presumably the puppet masters driving this whole scam) is too dumb to do this or that's really not how the numbers work out. Either that or all currencies everywhere are inflating at roughly the same rate without feeding back into wages anywhere, but that would be a really interesting macroeconomic state of affairs.
The intersection of "smart enough to be careful when planning your assassination" and "dumb / crazy enough to try assassinating a sitting US President" is probably very, very small. No doubt it's dwarfed by several orders of magnitude by the set "dumb enough to threaten to kill the President."
Comapring data points between two workers has all sort sof problems as you point out. But having all of the salary information public for everybody would lead to a much better estimate of the market price overall, which would be a good thing. Employers make arguments based on their individual quantization error and statistical outliers, but the real benefit is in the aggregate information. Knowing not just what your neighbor makes but whate everybody everywhere (including competing companies) makes allows you to make much better career decisions, the individual details about the guy in the next cube notwithstanding.
If we made that change, there would be a massive reshuffiling right off the bat, probably followed by a return to stability. If you have an individual case where you needed to make a weird lowball offer to somebody, they'd have the information needed to decide whether to take it up front. If you needed to pay more for a skill that was in demand on a short timeline, that would probably reflected in total market prices. Instead of having to pay your current people more to match the new employee's rate, you'd probably already have lost your employees who were making less than the market rate. Which is probably a good thing as well, because there's clearly plenty of high value work for them to be doing.
Take a look at the BLS data instead of sampling single data points. Like most people, your estimate of inflation is based on your gut and not on real data. Think about it this way: Assuming you're spot-on about soda and the price is climbing rapidly, what percentage of the average American household income goes to buying soda? How would even a small change in the price of some other good affect the usefulness of that data point?
But by the iPad Gen 1 index, inflation is way down. By the gold index, we're in deflation. By the California water index, inflation is out of control. But those aren't really indexes. They're just prices. The BLS puts a *ton* of effort into creating an index of what people actually buy that's comparable from year to year. A single price randomly sampled just won't do it. If you ask the average American, they seem to think they spend 95% of their income on gasoline and milk. But they don't. They spend a moderate percentage of their income on gas and hardly any of it on milk. They probably don't remember the 1/10th of a refrigerator they "bought" this year due to depreciation, for example.
The reality is that the BLS numbers match up well with the independent billion price index and with inflation expectations built into the financial markets. The odds that shadowstats or some other bogus index is the "real" index are slim to none. If we were really getting hammered at the the rates that, say, shadowstats, would have us believe, we'd all be destitute and we'd have noticed at least some captial flows following the money.
That's even better. Knowing what the guy in the next cube over makes just makes for a short-term rejiggering of the local salary distribution. Knowing with fairly high precision what you could be making literally anywhere else actually makes for good market rates. GlassDoor is a good start, but the market for employees would be much more efficient if all of that information was public and accurate. It would feel "icky" to us for a few months because of our culture and it would piss off corporate management (generally the benficiary of the current inefficiency), but the long run equilbiriubm would be better.
The grand problem isn't about a comparison between two employees and trying to justify why their salaries aren't identical. The real issue is that all of the salary information being public would actually result in good price setting information for everybody. Once we got over the initial shock that everybody doesn't make exactly the same amount of money, employees would actually be able to look at the aggregates and the data for other companies and make rational decisions about what they're really worth. As it stands, people are just guessing and a lot of suboptimal pricing is happening.
This is something that never made sense to me about the legal profession. If an engineer designed a bridge that violated building codes, he'd be in a world of shit. If a pilot ingored FAA regulations or safety best practices, that would be the end of his career. A lawyer can write a contract that's 80% unenforceable nonsense, sell that service to his employer as good legal work, and walk away with a paycheck and a blessing from his guild.
The non-union version would be a little different, though. Employers would have to pay their underpaid good performers more, and if budget started to get tight as salaries normalized, they'd have to get rid of the drooling idiot who wasn't earning his keep. In the union version, it's hard to get rid of the unproductive idiot for anything short of murder. In a competitive workplace, it's just a question of total salary budget.
That's a problem with management giving realistic assessments and comparisons to their employees, though. If you think you're above average and you're not, your manager should be able to give you a good reason why you're not. Ultimately after the initial shake up (even if it involves people moving to other companies where they'll *still* likely be rated as below average), people will settle into reality. The thought experiments people do about this stuff seems to conflict with real experiments. The NBA doesn't collapse into chaos because the news reports it whenever LeBron James gets a new contract. Other players know that they're making less than he is and they probably generally know why. It's just part of the culture, and the equilbrium dosn't seem to be any differnt from a workplace where salaries are kept a closely guarded secret. It would just take time for our culture to shift.
It doesn't "have" to be perfect to be a working market. It just has to be perfect in order to be a completely idealized perfectly competitive free market on paper. No markets live up to the perfectly idealized theory, but they generally work. They just work less well as you move away from that ideal state. "I don't know what my competitor whill charge tomorrow" is a slight deviation from perfect information, but it's not really a big deal. "Nobody knows what anybody else is charging," is a major deviation that causes real problems.
It's a problem that applies really well to the medical market as well: You don't know nearly as well as your doctor what the actual value of his services are. You also often don't know what he'll charge you until after you've "bought" his services. And you sure as hell don't have a list of what other doctors would charge you. The end result is that the market is a total mess. Every step away from that problem and toward more perfect information (clearly published prices, more educated consumers, etc.) will improve outcomes, even if we never get to true "perfect information."
That could be read one of two ways:
1) Different students might need to be taught in different ways to learn the same set of skills.
2) Different students should come out of school with very different sets of skills.
I'm totally onboard with (1), but not so much with (2). "He can't add, but boy can he write!" Everybody is going to have natural aptitudes for certain things, but there's a basic level of core competence that everybody should be able to hit.
We're constantly hybridizing and producing all manner of crazy new plants where God knows how many genes are combined in totally new ways. There are tons of hybrids we're eating right now that have nowhere near "many lifetimes" worth of consumption, and there's no good reason to think that we'll do a better job of predicting all of their subtle properties (especially "long term" safety) any better than if we take a known organism and add one carefully selected gene to it.
I'm in favor of good information on food labels, but I also see the complaint from the GMO producers. I think they're worried about something like this:
GMO company: There's no need to label these things. They're perfectly safe.
Anti-GMO activist: Why do you hate transparency? If it's perfectly safe, there's no reason not to label them.
[Time passes. Labels mandated.]
Anti-GMO activist: If GMOs are so safe, why is labeling them mandatory?
Consumers: Hey! That's a good point!
Same here. We think and solve problems similarly, so there isn't much "talking past each other" when we disagree. If a problem can be reasoned through, we usually come to agreement. If it's squishier, we both sort of recognize that it's probably not worth fighting to the death over. Money and worry about unemployment is never a problem, so we don't fight over it or nitpick how it's spent (in fact, we each have a mix of private and shared accounts after 10 years of marriage, and we each handle a subset of the household financees without much oversight from the other). Beyond that I suppose everything else is pretty normal, but those are two big issues that cause a lot of relationship trouble that work out to a pretty strong built-in advantage for us.
We're both pretty mellow people who don't usually get wound up about stuff that's not really important, but I'm not sure if that's just who we are or if it's associated with how we were educated and how we work. There seem to be certain types of people who genuinely get bored and need a certain amount of drama in their human interactions and will create it if it's not there. Those people seem to be less common in engineering, although I don't have a lot to quantify or support that. Those people are usually not fun to be around and seem like they'd be a nightmare to be married to.
I'm all for a massive reduction in police militarization and very nearly eliminating SWAT teams given how rarely they're used for what they were originally intended for. That being said, the problem with using this issue as a lever is that there will always be some fringe situation that calls for a swift armed reaction of some sort. These kids can literally say whatever they want, so there's always going to be a way to provoke a dangerous response as long as there's any conceivable situation that warrants that response. This is just one of those cases where there should be a panic button. It should be really difficult to hit and you should punish the living shit out of people who treat the panic button like it's a toy.
Let me step back a bit and rephrase my earlier question. There are two key price points: There's the upper limit, which is the amount of money you make for your employer. They wouldn't pay you a dollar more than that because they'd just be losing a dollar on the net, and they'd be indifferent between hiring you and not hiring you if it equals that amount because the net value to them is $0. Then there's the minimum amount you'd work for. You wouldn't work for less than that, by definition. So now there's a big range between those two numbers.
The salary you land on in that range reflects a split of that surplus value between you and your employer. You seem to think that you getting some of it while your employer keeps some of it is reasonable. You also seem to think that you getting none of it while your employer keeps all of it is reasonable. But somewhere along the line, you seem to think that you taking "too much" of it becomes unreasonable. I'm wondering what you're basing that on and how you decide when "too much" has happened. Why the asymmetry in favor of the employer, and why the certainty that Shearer crossed it somewhere between $300K and $450K per episode?
So if your employer cut you back to that base requirement, you wouldn't consider leaving and using your time some other way? Or at least work fewer hours?
Imagine you had enough money to live your current standard of living forever without working any more. Would that change your base requirement? You have a set of choices with respect to what to do with your time. One option would be not to work at all and to spend time with friends and family and enriching hobbies. Assuming that was an option to you, would the base pay requirement to get you to give that time up and go to work every day go up, or would it stay the same regardless of whether you needed the money or not?
I ask because Shearer has tons of money. A paycheck that would rent him a cheap apartment and a daily ration of Top Ramen would not affect his standard of living at all. He wouldn't notice either way. So given that he's probably a normal person with friends and family and hobbies, he'd be crazy to go to work at a job that only pays enough for a cheap apartment and a daily ration of Top Ramen. He's better off enjoying the days he has left doing something more enriching. Being "grateful" that an employer offered him a subsistence wage in exchange for losing that time wouldn't be rational at all.
If you had gotten paid a larger chunk of that money, would it have been morally wrong somehow?
Are you sure you deserved your whole salary as a software developer? I bet you'd have done it for less than you were paid, so why did you take that money? What are you really worth, and how do you know?
Becuse he's earning his employers shittons more than $6M per year by doing that work? Would the morally correct outcome be for him to cut his salary by, say, $3M so the owners of the company can pocket $3M more?
Monopoly or not, companies set their prices at the spot that they think maximizes their profits. They can't "pass on" 100% of a new cost by changing the price because that knocks them off of the spot they optimized for. They do lose something in the process. If all it took for them to grab an extra $229K in profits was to raise their prices, they'd have done it already.
White males are just about the easiest faces to categorize. They tend to have short hair that doesn't obscure facial features or create oddball shapes that confuse the classifiers. Their skin tone makes photographing them and finding edges extracting features easier than it is with darker skinned people. White people have a greater variety of eye colors that can be used to distinguish among them. "White guy face" is just about the optimal case for this problem. If I had to come up with a worst case that was also a photo of a fairly "common" person, I'd go with "dark skinned, brown eyed person wtih long hair and facial hair." That's a pretty clean sweep of all of the variables that make this a hard problem.
How good does the cutting edge of object recognition need to be before it's not "poorly tested" anymore, especially when it's for a silly photo app and not a medical or military application? I never hear this type of thing from people who have actually had to solve these types of problems. The reality is that objects are going to be confused with other objects. Lots of them, once we're talking about hundreds of millions or billions of samples. Some cases will fail with great regularity and patterns. The unfortunate fact here was that the pattern happened to coincidentally have really embarrassing cultural connotations.
This is one of the things I don't miss about working in machine vision. We'd run our algorithm over a zillion images and it would correctly handle all of them save a small handful and that small handful would be filed as bugs. OK, maybe we'll be able to handle that small handful at the expense of a smaller handful next time around. But the pass/fail criteria for the tool is in its overall results, not in the outliers.