Fortunately, it's fairly rare for the Supreme Court to decrease the scope of a prior seminal decision in order to reduce rights. Even the conservative members tend to enhance rights through judicial activism. Perhaps this will change, but Heller can't be underestimated -- I think it will turn out to be like Brown and Roe or similar in retrospect. The judicial slicing and dicing will be tempered by the realization that the same logic that applies to trying to limit Heller and McDonald can be applied to Miranda, Roe, and Brown (et al) and that will be a limitation on "living document" interpretation.
I assume you also believe that those who believe in respecting the other parts of the United States Constitution (for example, disallowing police from searching your home whenever they want for any or no reason without a warrant) are also "rabidly anti-government and completely reality-proof"?
Allowing police to search anyone, anywhere for any or no reason without notice or without a warrant would likely decrease crime significantly. That's certainly an admirable goal and, obviously, only a criminal who had something to hide would object. RIght?
And, after all, since we trust police with guns, we can surely trust them with less lethal tools like the ability to search freely.
You're statement is incorrect with respect to Federal laws (I don't know which, if any, states have stricter laws than the Feds on manufacturing firearms).
You might be correct if Lumpy had said that his friend sold or distributed the AK47 he made, but he didn't say that.
Check this out for more info (there are some restrictions, so read more than the portion I quoted):
For your information, per provisions of the Gun Control Act (GCA) of 1968, 18 U.S.C. Chapter 44, an unlicensed individual may make a “firearm” as defined in the GCA for his own personal use, but not for sale or distribution.
Which is, of course, only possible in the United States if the United States Constitution is first amended to nullify the Second Amendment. That only takes the approval of 3/4 of the State Legislatures.
The "amending" step is much harder than the act of making gun ownership illegal (which, itself, would be extremely difficult).
So, it's rather like if the police found a special car with very strong windows and combination locks. They have strong evidence that it's got a lot of heroin in it and want to get inside it to search it and have a warrant to do so but can't get it open.
They think, but don't have much evidence to support that belief, that you had unrestricted access to the car interior and therefore have the combination and can open the door for them.
What this ruling says is that they can't compel you to product the combination because then you would be being forced to reveal that you did, in fact, have the combination and, hence, access to the inside of the vehicle which would be incriminating given the contents of the car.
If, however, they found a surveillance video that showed you opening the door of the car using the combination you could then be compelled to provide the combination as that would not reveal, for the first time, that you actually had access to the interior of the car.
[...] A website surely has tacit consent to access the website or I'm in deep trouble, because I've been accessing information from millions of websites!
If a local mom-and-pop store lets you enter their premises to shop, do you think that also gives you the right to overturn the shelves and spray paint graffiti on the walls? After all, a mom-and-pop store isn't very significant in any (even tiny) country with a GDP over a billion USD and low wage employees can reverse the damage for only a few thousand dollars (less than will be spent to fix and harden a web site) so it's a minor crime. Really not a problem -- nothing a society would want to discourage (presumably even if every store suffered such an assault every couple hours -- after all, each individual one is minor). Right?
It also helps identify areas to design early (the "low confidence" areas) so the estimate can be refined, but more importantly, so unexpected architectural changes (most of which lurk in these "low confidence" areas) can be addressed while there's still time to address them while not in panic mode, while feature content can still be adjusted (eliminating a feature you've already coded and tested doesn't give you back any time and rarely saves you much future time), and while fewer customer promises have been made and affirmed.
Although, I find that less experienced developers often have a poor sense of which areas are low confidence (mostly because they don't think of as many "known unknowns" because they have never had their faces rubbed into them) so caution is advised.
The difficulty in this approach is that by the time you've gotten everything down to that level of detail, most of the risky and intellectual part of the work is done. Most of the rest can be done by cheap scalable coding monkeys. So you're now able to tell Marketing:
Recall that project you asked for a time and effort estimate on six months ago, well the answer is that it will be done nine months from when you asked (i.e., three months from now) and will require twelve person-years (of which we have already spent nine).
Marketing, of course, responds:
Huh, oh, we forgot about that - don't waste any more time on that because that idea was so 2012. But, since you're here, could you give me an elapsed time and effort estimate for $MY_LATEST_BRAIN_FART? By the way, what have you guys been doing the past six months?
Surprisingly... Things that you've done before, like brushing your teeth, are easy to estimate. Things that no one has done before are hard to estimate. The more interesting something is to the customer, the more likely it's of the latter class.
The solution seems to be "don't commit to a schedule longer than a sprint (even if that's only a week) and you won't be far off on the average".
Of course, this doesn't work so well with customers. A giant customer who is considering kicking your product out the door and replacing it with a competitor "if you don't get feature X in" wants to know when he can expect feature X. This is often easy for seemingly small projects (add a new style sheet), but not so much for "hard" (many tens of person years of development or more) which are relatively indivisible (a distributed system that only corrupts data in 20% of the cases for which code hasn't yet been written is about as useless as one that does so 100% of the time). In the hard projects, if you tell them:
Oh, we really have no idea, but I can confidently state that we will have these three small work items done a week from now and eventually expect we will have something done -- we will let you know the delivery date a week before we're ready to send you the software.
their answer will be simple - "Thanks for the information, we are cancelling our maintenance contract and won't be using your system anymore. Please give me your vendor badge which we have just deactivated anyway."
Did you miss that if you eat them all at once, they will stick together in one clump and therefore none would be in an "adjacent track of intestine". Although, I would think that just having a, effectively solid, chunk of indigestible material the size of several buckyballs may be a problem in of itself.
I was clearly referring to incompetent developers not being worth even a salary of $0. That sort of labor (which there is plenty of and which can quickly be created because anyone with an IQ over 90 can get a CS degree from somewhere and claim to be a "developer") has no (or negative) value to me. Skilled developers are valuable and that's what I try to hire (but, the vast majority of them are foreign).
So, no contradiction.
Hopefully you are not a developer because, obviously, you couldn't be in the "skilled" class since your reading comprehension problems would render you unable to even read a bug report accurately.
There are no "poor salaries" for competent developers. It's an extremely well paid profession - especially in tech centers such as the San Francisco Bay Area.
I care about the size of the pool because that's an important factor in determining where the United States will rank in the world in technical innovation and other important ways fifty years from now. I'd like to fix the US education system and culture so we mint more competent STEM grads. Unfortunately, this is hard and would, optimistically, take at least two full generations to accomplish. As well, the effects of doing this would not be noticed for 20 years even if it were magically implemented overnight as the problems start very early in childhood and can't be recovered from in most cases once the die is cast.
We do, of course, pay to hire the people we need -- it just turns out the pool of qualified candidates are about 95% foreign and started working in the US on a work visa of some sort (of course, many are now citizens or hold a Green Card). Imagine if not a single person who started working in the US on a work visa (including those who are now citizens) were in the US -- in that environment, very little software development COULD take place in the United States.
As a country, if we can't make as many great milkshakes as we need, we should import great milkshakes from other countries. We are drinking India's and China's great milkshakes -- that's what the whole H-1B program is about and it's good national policy (although, it should be expanded).
First, you WANT your manager to be someone who could return to "hacking" (a ridiculous term when dealing with Enterprise software where I have worked for 30 years) to remain a manager, you (and the shareholders whether they realize it or not) because you want them to be dealing with the corporate BS with a technical passion and using logic to set the PHBs back on their heels.
Few, albeit some, who have the native twisted brain to do great development choose to pursue a degree in Linguistics or Ancient Greek Literature just because of pay (degrees in Linguistics and Ancient Greek Literature are not known for great salaries and upsides due to stock options).
People who "stay in grad school" simply delay their enrtry into the developer pool by a few years AND, most of the time, are not really good developers (else, their passion for producing productive solutions would have overridden their fear of leaving academia).
I filter good from bad aggressively - 95% of the resumes I see don't even get a response to the headhunter/applicant. 98% don't get more than an email saying "not a match". However, I demand direct access to the stream, as much as HR hates losing control, I don't want to lose a good developer who may not know C++ but understands concurrency having been the key developer on some proprietary system I've (nor HR) has ever heard of. Only a tiny percentage of developers "get" systems development - I understand it, Windows GUI APIs let fresh-outs think they "understand" concurrency and multi-threading -- they rarely do.
simple economics suggests that they should offer more money
False. One can not mint truly skilled developers simply by paying them more - there's a pretty fixed inelastic supply legally able to work in the United States. There is, of course, a very elastic supply of incompetent developers -- but hiring one often, even at a salary of $0, costs more than leaving the position open.
I've read many thousands of resumes, done many hundreds of phone screens, and conducted hundreds of interviews over the years for developer positions. I've hired a bunch of developers, some of whom didn't work out (it's usually pretty obvious within a week or two that a hire was a mistake). I can't clone a good developer, I take them from a fixed pool and that pool isn't going to get bigger in the United States until we fix our lax culture and fix our educational system.
Okay. Since you want to make this personal. No, you're a fool.
MIT's open policy was simply a convenient exception to most institutions. However, the risk of the open policy interfering with productive use of the network has now, in the judgement of adults, exceeded the value of letting anyone run a child porn service (or similar, including DDOS attacks) on/from MIT's network. Early mass produced automobiles didn't have door locks or ignition locks - do you expect to have a door lock on a new car you buy? Time moves on.
Serious students who want to develop whatever they want to will simply set up N virtual machines on their laptop on a local virtual network to do whatever they need to do. If they want to expose it to the world, they will either apply for the "opt out" option with MIT or just use AWS or something like that to open it up to the broader world and end up launching the next Google or Facebook. It's not 1995 anymore - grow up - automobiles have door locks now.
Would we say that because MIT locks some of the doors to some of their rooms some of the time that the thieves and burglars have won long ago? Would we say that MIT "caved" to the thieves and burglars?
Re:Collateralized vs Non-Collateralized Loans
on
Let Them Eat Teslas
·
· Score: 1
At a minimum, we must "track" classes in the U.S.
In most public elementary schools (and for some purposes/levels, middle schools as well) in the area I live, every kid is pretty much assigned randomly to classrooms (some, who are behind, get additional "special" help in additional classes or sessions -- but it's generally not very effective unless there's an obvious issue like English as a second language).
So a fifth grade teacher has 25 students in the room ranging from the worst to the best in the school. The teacher has to teach "to the middle" without completely losing the lower kids. As a result, the "middle" students do okay, but are not challenged; the best students are bored (the teachers then often give them other advanced work, but the pace of the class is still MUCH too slow for them and sometimes they begin to goof off due to boredom and a lack of understanding that there ARE more challenging problems to be solving and that they should be solving if they can be considered to be "excelling"); and the worst students learn little and are demoralized. It really doesn't work very well for anyone.
The 'selecting a path in fifth grade' obviously needs some adaptation. However, I wonder if you and your parents knew that you would have been put in a path after, say, fifth grade, do you think that might have changed the motivation a bit? Most smart people seem to respond to concrete expectations and consequences pretty well but there are few set in elementary school.
Re:Collateralized vs Non-Collateralized Loans
on
Let Them Eat Teslas
·
· Score: 1
Transitioning between the levels in the German system has been made easier than it once was, true and such a program in the U.S. should have the same feature. However, as you say, it's based on PERFORMANCE and preference. The case you describe of course sounds like someone who decided not to go to college even though he was in the college track and (presumably) could have -- of course, no system should force someone to go to college.
Re:Collateralized vs Non-Collateralized Loans
on
Let Them Eat Teslas
·
· Score: 1
Make it the greater of x% of the first y thousand hours of W2 reported work (add hours to the W2 for the sake of this program) and AGI for the first z years where x, y and z are picked by the loan providers for each applicant. That would take care of the part timers.
Re:also need to cut fluff and filler from Educatio
on
Let Them Eat Teslas
·
· Score: 1
Did someone forget that it's April 3, not April 1?
Re:Collateralized vs Non-Collateralized Loans
on
Let Them Eat Teslas
·
· Score: 5, Insightful
Unfortunately, in the U.S., I fear this would result in a situation where "everyone gets an associates or technical degree", and these programs would end up getting dumbed down so "everyone gets an associates or technical degree" in the interest of "fairness, diversity, and inclusiveness" - just like our K-12 system has. Then, the two year degree would be worth what a HS degree is now - not much.
I think the U.S. should be seriously considering something more like the German system. Put students on one of a few tracks early (maybe fifth grade?). One of these would be a strong academic track which is intended to end up with at least solid STEM MSc or a PhD in one of the "softer and in less demand" disciplines (such as Ancient History). At the other end of the spectrum would be trade oriented curriculum with students learning skills necessary for basic life (some accounting, enough math to figure out interest, basic understanding of government, basic writing skills) and skills specific to a group of trades and including apprenticeships in HS.
If you've ever worked with fifth graders, it's pretty obvious which ones will never be on the academic the track or the one just "below" it. If, in spite of being exposed multiple times to it, they haven't figured out how to compute 6.9 × 0.042 or 42.42 ÷ 0.1, they are extremely unlikely to catch up and even survive Algebra 2 a few years later. However, generally students should be given the benefit of the doubt and put into a higher track if they are "on the bubble" and then be aggressively reevaluated every six months to determine if they should be moved to a less academically demanding track. The requirements and expectations of the tracks should not be adjusted to match the students, the students should be placed in the track that match their motivation, intelligence, and interests.
The case is more compelling if there is evidence of systemic prejudice -- but it's not required. A single case is enough.
Consider if you (having not gone to training!) have been heard saying "I wish the married employees would work as hard as the single ones" and a few weeks later a manager fires a married employee because they don't seem to mesh very well with the team. One of the justifications is that you found the fired employee hard to work with.
In my state (an "at will" state) the employee will have a pretty decent case if they claim they were fired, at least in part, because of their marital status and the manager doesn't have much other objective evidence that their stated reason was the real reason. The employee may not be successful if it goes to jury trial - but even if not, it will be expensive for the company (if you dislike sitting in a one hour training session once a year, you would really hate sitting in depositions for hours). The company may just end up paying a "cost of defense" (and a bit more if the lawyer is any good) settlement.
"At will" is not blanket immunity when dealing with discrimination against protected classes.
Slashdot Mirror
Fortunately, it's fairly rare for the Supreme Court to decrease the scope of a prior seminal decision in order to reduce rights. Even the conservative members tend to enhance rights through judicial activism. Perhaps this will change, but Heller can't be underestimated -- I think it will turn out to be like Brown and Roe or similar in retrospect. The judicial slicing and dicing will be tempered by the realization that the same logic that applies to trying to limit Heller and McDonald can be applied to Miranda, Roe, and Brown (et al) and that will be a limitation on "living document" interpretation.
Agreed -- i was assuming that he was talking about something that most people could purchase legally.
I assume you also believe that those who believe in respecting the other parts of the United States Constitution (for example, disallowing police from searching your home whenever they want for any or no reason without a warrant) are also "rabidly anti-government and completely reality-proof"?
Allowing police to search anyone, anywhere for any or no reason without notice or without a warrant would likely decrease crime significantly. That's certainly an admirable goal and, obviously, only a criminal who had something to hide would object. RIght?
And, after all, since we trust police with guns, we can surely trust them with less lethal tools like the ability to search freely.
You're statement is incorrect with respect to Federal laws (I don't know which, if any, states have stricter laws than the Feds on manufacturing firearms).
You might be correct if Lumpy had said that his friend sold or distributed the AK47 he made, but he didn't say that.
Check this out for more info (there are some restrictions, so read more than the portion I quoted):
Which is, of course, only possible in the United States if the United States Constitution is first amended to nullify the Second Amendment. That only takes the approval of 3/4 of the State Legislatures.
The "amending" step is much harder than the act of making gun ownership illegal (which, itself, would be extremely difficult).
I'd probably just say "No" or "Why don't you use your card?" -- how the employer responds would inform my decision on what to do next.
So, it's rather like if the police found a special car with very strong windows and combination locks. They have strong evidence that it's got a lot of heroin in it and want to get inside it to search it and have a warrant to do so but can't get it open.
They think, but don't have much evidence to support that belief, that you had unrestricted access to the car interior and therefore have the combination and can open the door for them.
What this ruling says is that they can't compel you to product the combination because then you would be being forced to reveal that you did, in fact, have the combination and, hence, access to the inside of the vehicle which would be incriminating given the contents of the car.
If, however, they found a surveillance video that showed you opening the door of the car using the combination you could then be compelled to provide the combination as that would not reveal, for the first time, that you actually had access to the interior of the car.
Is that correct?
If a local mom-and-pop store lets you enter their premises to shop, do you think that also gives you the right to overturn the shelves and spray paint graffiti on the walls? After all, a mom-and-pop store isn't very significant in any (even tiny) country with a GDP over a billion USD and low wage employees can reverse the damage for only a few thousand dollars (less than will be spent to fix and harden a web site) so it's a minor crime. Really not a problem -- nothing a society would want to discourage (presumably even if every store suffered such an assault every couple hours -- after all, each individual one is minor). Right?
I've done something similar to this as well.
It also helps identify areas to design early (the "low confidence" areas) so the estimate can be refined, but more importantly, so unexpected architectural changes (most of which lurk in these "low confidence" areas) can be addressed while there's still time to address them while not in panic mode, while feature content can still be adjusted (eliminating a feature you've already coded and tested doesn't give you back any time and rarely saves you much future time), and while fewer customer promises have been made and affirmed.
Although, I find that less experienced developers often have a poor sense of which areas are low confidence (mostly because they don't think of as many "known unknowns" because they have never had their faces rubbed into them) so caution is advised.
The difficulty in this approach is that by the time you've gotten everything down to that level of detail, most of the risky and intellectual part of the work is done. Most of the rest can be done by cheap scalable coding monkeys. So you're now able to tell Marketing:
Marketing, of course, responds:
Surprisingly... Things that you've done before, like brushing your teeth, are easy to estimate. Things that no one has done before are hard to estimate. The more interesting something is to the customer, the more likely it's of the latter class.
Pretty good summary.
The solution seems to be "don't commit to a schedule longer than a sprint (even if that's only a week) and you won't be far off on the average".
Of course, this doesn't work so well with customers. A giant customer who is considering kicking your product out the door and replacing it with a competitor "if you don't get feature X in" wants to know when he can expect feature X. This is often easy for seemingly small projects (add a new style sheet), but not so much for "hard" (many tens of person years of development or more) which are relatively indivisible (a distributed system that only corrupts data in 20% of the cases for which code hasn't yet been written is about as useless as one that does so 100% of the time). In the hard projects, if you tell them:
their answer will be simple - "Thanks for the information, we are cancelling our maintenance contract and won't be using your system anymore. Please give me your vendor badge which we have just deactivated anyway."
Did you miss that if you eat them all at once, they will stick together in one clump and therefore none would be in an "adjacent track of intestine". Although, I would think that just having a, effectively solid, chunk of indigestible material the size of several buckyballs may be a problem in of itself.
If you find it necessary to zip up your fly while driving, you may have been engaged in a dangerously distracting activity recently.
I was clearly referring to incompetent developers not being worth even a salary of $0. That sort of labor (which there is plenty of and which can quickly be created because anyone with an IQ over 90 can get a CS degree from somewhere and claim to be a "developer") has no (or negative) value to me. Skilled developers are valuable and that's what I try to hire (but, the vast majority of them are foreign).
So, no contradiction.
Hopefully you are not a developer because, obviously, you couldn't be in the "skilled" class since your reading comprehension problems would render you unable to even read a bug report accurately.
There are no "poor salaries" for competent developers. It's an extremely well paid profession - especially in tech centers such as the San Francisco Bay Area.
I care about the size of the pool because that's an important factor in determining where the United States will rank in the world in technical innovation and other important ways fifty years from now. I'd like to fix the US education system and culture so we mint more competent STEM grads. Unfortunately, this is hard and would, optimistically, take at least two full generations to accomplish. As well, the effects of doing this would not be noticed for 20 years even if it were magically implemented overnight as the problems start very early in childhood and can't be recovered from in most cases once the die is cast.
We do, of course, pay to hire the people we need -- it just turns out the pool of qualified candidates are about 95% foreign and started working in the US on a work visa of some sort (of course, many are now citizens or hold a Green Card). Imagine if not a single person who started working in the US on a work visa (including those who are now citizens) were in the US -- in that environment, very little software development COULD take place in the United States.
As a country, if we can't make as many great milkshakes as we need, we should import great milkshakes from other countries. We are drinking India's and China's great milkshakes -- that's what the whole H-1B program is about and it's good national policy (although, it should be expanded).
First, you WANT your manager to be someone who could return to "hacking" (a ridiculous term when dealing with Enterprise software where I have worked for 30 years) to remain a manager, you (and the shareholders whether they realize it or not) because you want them to be dealing with the corporate BS with a technical passion and using logic to set the PHBs back on their heels.
Few, albeit some, who have the native twisted brain to do great development choose to pursue a degree in Linguistics or Ancient Greek Literature just because of pay (degrees in Linguistics and Ancient Greek Literature are not known for great salaries and upsides due to stock options).
People who "stay in grad school" simply delay their enrtry into the developer pool by a few years AND, most of the time, are not really good developers (else, their passion for producing productive solutions would have overridden their fear of leaving academia).
I filter good from bad aggressively - 95% of the resumes I see don't even get a response to the headhunter/applicant. 98% don't get more than an email saying "not a match". However, I demand direct access to the stream, as much as HR hates losing control, I don't want to lose a good developer who may not know C++ but understands concurrency having been the key developer on some proprietary system I've (nor HR) has ever heard of. Only a tiny percentage of developers "get" systems development - I understand it, Windows GUI APIs let fresh-outs think they "understand" concurrency and multi-threading -- they rarely do.
False. One can not mint truly skilled developers simply by paying them more - there's a pretty fixed inelastic supply legally able to work in the United States. There is, of course, a very elastic supply of incompetent developers -- but hiring one often, even at a salary of $0, costs more than leaving the position open.
I've read many thousands of resumes, done many hundreds of phone screens, and conducted hundreds of interviews over the years for developer positions. I've hired a bunch of developers, some of whom didn't work out (it's usually pretty obvious within a week or two that a hire was a mistake). I can't clone a good developer, I take them from a fixed pool and that pool isn't going to get bigger in the United States until we fix our lax culture and fix our educational system.
Okay. Since you want to make this personal. No, you're a fool.
MIT's open policy was simply a convenient exception to most institutions. However, the risk of the open policy interfering with productive use of the network has now, in the judgement of adults, exceeded the value of letting anyone run a child porn service (or similar, including DDOS attacks) on/from MIT's network. Early mass produced automobiles didn't have door locks or ignition locks - do you expect to have a door lock on a new car you buy? Time moves on.
Serious students who want to develop whatever they want to will simply set up N virtual machines on their laptop on a local virtual network to do whatever they need to do. If they want to expose it to the world, they will either apply for the "opt out" option with MIT or just use AWS or something like that to open it up to the broader world and end up launching the next Google or Facebook. It's not 1995 anymore - grow up - automobiles have door locks now.
Would we say that because MIT locks some of the doors to some of their rooms some of the time that the thieves and burglars have won long ago? Would we say that MIT "caved" to the thieves and burglars?
At a minimum, we must "track" classes in the U.S.
In most public elementary schools (and for some purposes/levels, middle schools as well) in the area I live, every kid is pretty much assigned randomly to classrooms (some, who are behind, get additional "special" help in additional classes or sessions -- but it's generally not very effective unless there's an obvious issue like English as a second language).
So a fifth grade teacher has 25 students in the room ranging from the worst to the best in the school. The teacher has to teach "to the middle" without completely losing the lower kids. As a result, the "middle" students do okay, but are not challenged; the best students are bored (the teachers then often give them other advanced work, but the pace of the class is still MUCH too slow for them and sometimes they begin to goof off due to boredom and a lack of understanding that there ARE more challenging problems to be solving and that they should be solving if they can be considered to be "excelling"); and the worst students learn little and are demoralized. It really doesn't work very well for anyone.
The 'selecting a path in fifth grade' obviously needs some adaptation. However, I wonder if you and your parents knew that you would have been put in a path after, say, fifth grade, do you think that might have changed the motivation a bit? Most smart people seem to respond to concrete expectations and consequences pretty well but there are few set in elementary school.
Transitioning between the levels in the German system has been made easier than it once was, true and such a program in the U.S. should have the same feature. However, as you say, it's based on PERFORMANCE and preference. The case you describe of course sounds like someone who decided not to go to college even though he was in the college track and (presumably) could have -- of course, no system should force someone to go to college.
Make it the greater of x% of the first y thousand hours of W2 reported work (add hours to the W2 for the sake of this program) and AGI for the first z years where x, y and z are picked by the loan providers for each applicant. That would take care of the part timers.
Did someone forget that it's April 3, not April 1?
Unfortunately, in the U.S., I fear this would result in a situation where "everyone gets an associates or technical degree", and these programs would end up getting dumbed down so "everyone gets an associates or technical degree" in the interest of "fairness, diversity, and inclusiveness" - just like our K-12 system has. Then, the two year degree would be worth what a HS degree is now - not much.
I think the U.S. should be seriously considering something more like the German system. Put students on one of a few tracks early (maybe fifth grade?). One of these would be a strong academic track which is intended to end up with at least solid STEM MSc or a PhD in one of the "softer and in less demand" disciplines (such as Ancient History). At the other end of the spectrum would be trade oriented curriculum with students learning skills necessary for basic life (some accounting, enough math to figure out interest, basic understanding of government, basic writing skills) and skills specific to a group of trades and including apprenticeships in HS.
If you've ever worked with fifth graders, it's pretty obvious which ones will never be on the academic the track or the one just "below" it. If, in spite of being exposed multiple times to it, they haven't figured out how to compute 6.9 × 0.042 or 42.42 ÷ 0.1, they are extremely unlikely to catch up and even survive Algebra 2 a few years later. However, generally students should be given the benefit of the doubt and put into a higher track if they are "on the bubble" and then be aggressively reevaluated every six months to determine if they should be moved to a less academically demanding track. The requirements and expectations of the tracks should not be adjusted to match the students, the students should be placed in the track that match their motivation, intelligence, and interests.
The case is more compelling if there is evidence of systemic prejudice -- but it's not required. A single case is enough.
Consider if you (having not gone to training!) have been heard saying "I wish the married employees would work as hard as the single ones" and a few weeks later a manager fires a married employee because they don't seem to mesh very well with the team. One of the justifications is that you found the fired employee hard to work with.
In my state (an "at will" state) the employee will have a pretty decent case if they claim they were fired, at least in part, because of their marital status and the manager doesn't have much other objective evidence that their stated reason was the real reason. The employee may not be successful if it goes to jury trial - but even if not, it will be expensive for the company (if you dislike sitting in a one hour training session once a year, you would really hate sitting in depositions for hours). The company may just end up paying a "cost of defense" (and a bit more if the lawyer is any good) settlement.
"At will" is not blanket immunity when dealing with discrimination against protected classes.