"The company said Wednesday that it remained committed to creating as many as 13,000 jobs in Wisconsin, and continued to "actively consider opportunities" involving flat-screen technology. But it said it was also "examining ways for Wisconsin's knowledge workers to promote research and development."
The change is they are backing off the reduced 1000 worker plan, going back to shooting for 13,000 and planning to add R&D into the mix.
"tax breaks that a 1000 employee company could have never negotiated on their own"
That is also bit of a stretch. Microsoft was a massive global monopoly at 100 workers. They certainly could have negotiated a 3 billion dollar deal with a state. A few thousand R&D jobs on the roster could kickstart Wisconsin being seriously competitive in the modern world by instigating a revolution in the educated workforce and reduce incentive for their best and brightest to flee to other states like California.
Do everyone a favor and leave your politics on your nightstand, they cloud your judgement.
That's a chicken an egg problem. If you build it, they will come.
It is too expensive to source locally, that is the reason for the subsidy. It isn't as if there aren't plenty of raw resources in the US that could be refined. Those industries have just diminished and stagnated. The idea here was to increase pressure to develop them again. The end result being the same supply chain but in the US instead of Asia.
$230k in potential subsidies per worker, new tariffs on Chinese imports and a company that needs plenty of R&D.
With this announcement Foxconn states they want to go from a plan to employ 1000 blue collar workers to a plan to utilize the plant for R&D. Much higher paid and higher education employees. This would let Foxconn make the most of the subsidy, avoid the tariff issues, and make a big push to a higher class of worker in Wisconsin.
Yes, you could argue that anything Foxconn says should be treated with a heavy grain of salt at this point but consider for a moment that maybe Foxconn doesn't want to either rob the state of Wisconsin or engage in a bad business decision. This would be the direction they'd have to start going down. It really would be a win all around.
Now to part B, making sure it is actually natural born Americans staffing the place. Preferably Wisconsonians. Although Wisconsin's mostly white demographic will make you cringe Ami so I know you'll oppose that and force a racist demand they import people to avoid an uneven diversity profile.
"They only issue now is how much of the credits are dependent on performance by Foxconn."
If they actually mean what they are saying that isn't a problem. They are talking about white collar (as in lab coat) jobs instead of manufacturing jobs. This makes sense for Foxconn which needs a lot of R&D. Those staff will have better than the $55k/yr avg salary originally estimated and actually would be an upgrade for Wisconsin.
Granted that depends on follow through from Foxconn but there is a big carrot being offered by Wisconsin and this seems like a great way for them to make the most of it.
People aren't even reading this. Yes this is an announcement that they aren't going to fill manufacturing jobs. Now they are talking about R&D jobs. $230k/employee means it makes more sense to bring in employees that let them take advantage of the $230k/employee.
Is this going to help the blue collar workers of Wisconsin? Indirectly since there will be support and infrastructure jobs around this. But if they actually follow through it will help create (or drastically expand) a new class of educated worker and industry in Wisconsin.
Proportional representation in results should be the goal. If it's a purple state then the candidates should be purple. It doesn't matter if a D or R is elected, they should be representing the interests of the voters who went the other way in proportion.
"What you don't want is to keep people in minimum wage manufacturing jobs."
Minimum wage? Manufacturing and factory jobs paid about $20-35/hr in the 90's to early 00's. That's substantially more than the massive minimum wage increase people are pushing and no small number of them are skilled labor. That doesn't go very far in a large city but in many ways you live more comfortably on that income in the rural midwest than you do in on twice that in the city because the cost of living is so much lower.
We do need to stop charging so much for education but lets not get carried away. A moron is a moron and you just end up with a bunch of over-educated morons going down that path. We already see that in many fields. The idea isn't for everyone to succeed, the idea is for those with drive, ambition, and capability to succeed. I'd like to see a more even playing field amongst economic classes in that race so those people achieve equal success regardless of their parents success. That doesn't mean we want bottom feeders who can parrot a professor and look down on the "ignorant" with a sneer to succeed. In other words, the last thing we want is to be the new Europe.
That's what I think is mischaracterized here. People are talking about it like it's some big payout Wisconsin made to Foxconn and they can run off with. It's an opportunity, if Foxconn doesn't take advantage of it they won't actually get the money.
Also, what Foxconn is reconsidering according to TFS is their decision to employ only a 1000 workers. Now they are talking about moving R&D there rather than manufacturing. Why? They want the carrot Wisconsin offered. That will actually bring in higher salaries and revenues for the state. It just won't help with the blue collar workers that were supposed to be targeted.
It isn't like they wrote a check to Foxconn. There are no tax breaks, land, and infrastructure that needs to be given and therefore recouped if they aren't building. The state can always just take the land back.
None of that does them any good if they don't utilize it. Hell, Wisconsin can (and probably should) simply take it back and find someone else who wants the spot.
You are beating a strawman. I never claimed Python was difficult to read only that well written Perl can surpass it. Python is a perfectly functional and readable language, it's just a solution looking for a problem.
But since you are fishing for complaints, Python has plenty of issues and most of them are just arbitrary and inconsistent design. How about chapter 1, variables.
Remind me again, is this is a list, a tuple, or a dictionary?
ambiguous[1]
Oh right, you can't tell because they use the same braces outside declaration to avoid using as many characters.
What type does this give int(10)/int(3)?
Oh right, they arbitrarily changed the behavior between python versions introducing thousands of bugs should you upgrade.
How many elements does this return?
string[0:1]
Again, we have odd and arbitrary behavior, for some reason the first element specified is inclusive but the second is exclusive.
What does this give 2^4 how about this 2^8?
Not what one would normally expect since Python decided to break conventions on math symbols, again apparently just for the purpose of reducing commonly used characters.
They don't have to. They just have to them well enough.
"A major part of the problem is that we can't even foresee what tasks someone like a doctor might have to do, as they regularly encounter novel situations for which there aren't millions of documented cases to train a NN on."
That isn't really how it works. We focus on one thing at a time. At some point the low hanging fruit and individual tasks the AI and automation can perform can be stitched together to take over most of the work. Now you need an AI supervised by a couple doctors instead of 200 doctors. Eventually the cases where the AI falls down are few enough to just let those fail, after all the doctor fails sometimes as well.
Right now there are massive classes of surgery where the doctor basically performs the surgery by pressing a button to tell the machine to do the work. The need for a doctor is being artificially supported in the device. There is already software which has been tested alongside doctors in hospitals and was more successful at diagnosis.
Your logic regarding novel situations falls down because if they were commonly encountered they wouldn't be novel and the doctor who is actually good in a novel situation isn't typical.
Neither is hard to read if you bother to learn the language. Some badly written Perl can be hard to read, some well written Perl can be more legible than Python (which is all basically the same because of enforced whitespace).
If you find Perl incomprehensible you must not have ever really learned Perl. Just like if regular expressions are incomprehensible to you then you've never really learned them. Now try determining an associative array or hash with numeric keys isn't a regular array in Python when the access syntax is identical. Oh wait I meant list vs dictionary since industry standard naming isn't used in Python.
No less inexplicable than java itself but when you are right you are right.
There is essentially nothing Python does better than Perl other than utilize libraries, api's, and tutorials written by people inexplicably being taught python now rather than Perl. Perl is faster when written well, more comprehensive, more flexible, is more internally consistent, and with a style guideline (which can even be automatically applied for you) its easier to read.
My only explanation is that a bunch of ignorant people who thought Perl was dead because Perl 6 never took off and gained their impression of Perl legibility from script kiddies coding CGI are responsible for pushing out Python.
All that aside there is no denying reality and I'm starting to force myself to pull Python out of my bag in place of Perl these days. Despite contributors to cpan doing what they can to fill the holes it is just an uphill battle in this world of cloud tools and apis everywhere when nobody releases a Perl version of anything.
Anecdotally I can assure you that javascript is EXTREMELY popular, it is so popular in fact that entire classes of job description simply omit it as an assumed requirement just like the ability to handle a few windows systems is assumed when hiring a Unix guy.
Of course, not to mention technical jobs. People are always so naive on these things. They obsess over incredibly expensive robots that can perform some sort of manual labor when it is NN 'bots' replacing high paid jobs they should worry about. Why spend all that money cutting out your lowest paid jobs when you can spend it cutting out some of your highest paid ones?
"At least in the US there's a history of these kind of projects"
There is a pretty well established history for both actually. I'm certainly not claiming the US are the good guys vs Russia who are the bad guys, Putin, China, North Korea are all countries with a long public record of being bad actors and there is little to no credible dispute of that. People do dispute the US, especially on a US forum like this one. As seen by someone replying and trying to paint the suggestion as a crackpot conspiracy theory.
Putin's crimes are just far more thinly veiled with a history of poorly covered domestic murder and death squads. The US on the other hand has aired much of its historical dirty laundry and so far has convinced the public of the "this is how we used to do things back in the wild west of history" dialog. Why people just blindly accept this is the past and anyone suggesting the US government has continued with business as usual is crazy and paranoid I don't know. Maybe it's just some very effective propaganda.
I think you are confusing the API, the abstracted standard and documentation which defines the interface, with the coded definition in their implementation. It's understandable because there will certainly be a massive amount of overlap in a new clean room implementation using that abstracted standard.
This is like a new state coming up with a drivers license and another state claiming it infringes on their copyright because it contains a picture, a dob, and a hologram and they look strikingly similar to that states and interoperability doesn't apply because the image in the hologram is different. In reality the new state took written descriptions of the security measures used by other states and came up with their own but in practice any implementation of those standards is going to end up looking similar. There are only so many ways to implement a clear head shot and so many ways to display a DOB. The same is true here but to an even greater extent, given there are common coding styles and practices a good portion of it likely would end up being identical character for character.
"The supreme court already was asked to rule and that point, and they let the appellate court's ruling stand. "
The story we are commenting under is a new request that isn't settled and most of the circuits currently agree that API's can't be copyrighted including the one that matters.
"and there is no legal theory to distinguish APIs from any other part of code"
APIs aren't part of the code. You really shouldn't comment on topics you are ignorant of and really neither should courts but sadly we are stuck with whatever random nonsense they spew from ignorance.
Slashdot Mirror
"The company said Wednesday that it remained committed to creating as many as 13,000 jobs in Wisconsin, and continued to "actively consider opportunities" involving flat-screen technology. But it said it was also "examining ways for Wisconsin's knowledge workers to promote research and development."
The change is they are backing off the reduced 1000 worker plan, going back to shooting for 13,000 and planning to add R&D into the mix.
"tax breaks that a 1000 employee company could have never negotiated on their own"
That is also bit of a stretch. Microsoft was a massive global monopoly at 100 workers. They certainly could have negotiated a 3 billion dollar deal with a state. A few thousand R&D jobs on the roster could kickstart Wisconsin being seriously competitive in the modern world by instigating a revolution in the educated workforce and reduce incentive for their best and brightest to flee to other states like California.
Do everyone a favor and leave your politics on your nightstand, they cloud your judgement.
"I don't see any provision for clawbacks by the state or local officials."
They don't need one. State and local government can appropriate land at will if they decide they need it. How do you think roads get built?
Normally they have to pay "fair value" but they get to decide what that is. In a case like this they could simply call $1 fair value.
That's a chicken an egg problem. If you build it, they will come.
It is too expensive to source locally, that is the reason for the subsidy. It isn't as if there aren't plenty of raw resources in the US that could be refined. Those industries have just diminished and stagnated. The idea here was to increase pressure to develop them again. The end result being the same supply chain but in the US instead of Asia.
$230k in potential subsidies per worker, new tariffs on Chinese imports and a company that needs plenty of R&D.
With this announcement Foxconn states they want to go from a plan to employ 1000 blue collar workers to a plan to utilize the plant for R&D. Much higher paid and higher education employees. This would let Foxconn make the most of the subsidy, avoid the tariff issues, and make a big push to a higher class of worker in Wisconsin.
Yes, you could argue that anything Foxconn says should be treated with a heavy grain of salt at this point but consider for a moment that maybe Foxconn doesn't want to either rob the state of Wisconsin or engage in a bad business decision. This would be the direction they'd have to start going down. It really would be a win all around.
Now to part B, making sure it is actually natural born Americans staffing the place. Preferably Wisconsonians. Although Wisconsin's mostly white demographic will make you cringe Ami so I know you'll oppose that and force a racist demand they import people to avoid an uneven diversity profile.
"They only issue now is how much of the credits are dependent on performance by Foxconn."
If they actually mean what they are saying that isn't a problem. They are talking about white collar (as in lab coat) jobs instead of manufacturing jobs. This makes sense for Foxconn which needs a lot of R&D. Those staff will have better than the $55k/yr avg salary originally estimated and actually would be an upgrade for Wisconsin.
Granted that depends on follow through from Foxconn but there is a big carrot being offered by Wisconsin and this seems like a great way for them to make the most of it.
People aren't even reading this. Yes this is an announcement that they aren't going to fill manufacturing jobs. Now they are talking about R&D jobs. $230k/employee means it makes more sense to bring in employees that let them take advantage of the $230k/employee.
Is this going to help the blue collar workers of Wisconsin? Indirectly since there will be support and infrastructure jobs around this. But if they actually follow through it will help create (or drastically expand) a new class of educated worker and industry in Wisconsin.
"they would want to transfer some of the manufacturing knowledge and skill they developed at home"
Like what? Chinese manufacturing knowledge is entirely stolen from the US. We built their industry.
Proportional representation in results should be the goal. If it's a purple state then the candidates should be purple. It doesn't matter if a D or R is elected, they should be representing the interests of the voters who went the other way in proportion.
"What you don't want is to keep people in minimum wage manufacturing jobs."
Minimum wage? Manufacturing and factory jobs paid about $20-35/hr in the 90's to early 00's. That's substantially more than the massive minimum wage increase people are pushing and no small number of them are skilled labor. That doesn't go very far in a large city but in many ways you live more comfortably on that income in the rural midwest than you do in on twice that in the city because the cost of living is so much lower.
We do need to stop charging so much for education but lets not get carried away. A moron is a moron and you just end up with a bunch of over-educated morons going down that path. We already see that in many fields. The idea isn't for everyone to succeed, the idea is for those with drive, ambition, and capability to succeed. I'd like to see a more even playing field amongst economic classes in that race so those people achieve equal success regardless of their parents success. That doesn't mean we want bottom feeders who can parrot a professor and look down on the "ignorant" with a sneer to succeed. In other words, the last thing we want is to be the new Europe.
That's what I think is mischaracterized here. People are talking about it like it's some big payout Wisconsin made to Foxconn and they can run off with. It's an opportunity, if Foxconn doesn't take advantage of it they won't actually get the money.
Also, what Foxconn is reconsidering according to TFS is their decision to employ only a 1000 workers. Now they are talking about moving R&D there rather than manufacturing. Why? They want the carrot Wisconsin offered. That will actually bring in higher salaries and revenues for the state. It just won't help with the blue collar workers that were supposed to be targeted.
It isn't like they wrote a check to Foxconn. There are no tax breaks, land, and infrastructure that needs to be given and therefore recouped if they aren't building. The state can always just take the land back.
None of that does them any good if they don't utilize it. Hell, Wisconsin can (and probably should) simply take it back and find someone else who wants the spot.
If that is all you are going for you can just kick off infrastructure.
You are beating a strawman. I never claimed Python was difficult to read only that well written Perl can surpass it. Python is a perfectly functional and readable language, it's just a solution looking for a problem.
But since you are fishing for complaints, Python has plenty of issues and most of them are just arbitrary and inconsistent design. How about chapter 1, variables.
Remind me again, is this is a list, a tuple, or a dictionary?
ambiguous[1]
Oh right, you can't tell because they use the same braces outside declaration to avoid using as many characters.
What type does this give int(10)/int(3)?
Oh right, they arbitrarily changed the behavior between python versions introducing thousands of bugs should you upgrade.
How many elements does this return?
string[0:1]
Again, we have odd and arbitrary behavior, for some reason the first element specified is inclusive but the second is exclusive.
What does this give 2^4 how about this 2^8?
Not what one would normally expect since Python decided to break conventions on math symbols, again apparently just for the purpose of reducing commonly used characters.
As a German friend of mine once said about why he moved to the US. In Europe there is a left, a far left, and an even further left.
"They're very far from doing all tasks better."
They don't have to. They just have to them well enough.
"A major part of the problem is that we can't even foresee what tasks someone like a doctor might have to do, as they regularly encounter novel situations for which there aren't millions of documented cases to train a NN on."
That isn't really how it works. We focus on one thing at a time. At some point the low hanging fruit and individual tasks the AI and automation can perform can be stitched together to take over most of the work. Now you need an AI supervised by a couple doctors instead of 200 doctors. Eventually the cases where the AI falls down are few enough to just let those fail, after all the doctor fails sometimes as well.
Right now there are massive classes of surgery where the doctor basically performs the surgery by pressing a button to tell the machine to do the work. The need for a doctor is being artificially supported in the device. There is already software which has been tested alongside doctors in hospitals and was more successful at diagnosis.
Your logic regarding novel situations falls down because if they were commonly encountered they wouldn't be novel and the doctor who is actually good in a novel situation isn't typical.
"I find PERL incomprehensible ...
And you find Python hard to read ..."
Neither is hard to read if you bother to learn the language. Some badly written Perl can be hard to read, some well written Perl can be more legible than Python (which is all basically the same because of enforced whitespace).
If you find Perl incomprehensible you must not have ever really learned Perl. Just like if regular expressions are incomprehensible to you then you've never really learned them. Now try determining an associative array or hash with numeric keys isn't a regular array in Python when the access syntax is identical. Oh wait I meant list vs dictionary since industry standard naming isn't used in Python.
Interpreted languages are still programming languages. Javascript qualifies with NodeJS. Markup languages aren't for instructing computers at all.
Do things like shell script qualify? Of course not.
Many of them can actually be compiled as well.
No less inexplicable than java itself but when you are right you are right.
There is essentially nothing Python does better than Perl other than utilize libraries, api's, and tutorials written by people inexplicably being taught python now rather than Perl. Perl is faster when written well, more comprehensive, more flexible, is more internally consistent, and with a style guideline (which can even be automatically applied for you) its easier to read.
My only explanation is that a bunch of ignorant people who thought Perl was dead because Perl 6 never took off and gained their impression of Perl legibility from script kiddies coding CGI are responsible for pushing out Python.
All that aside there is no denying reality and I'm starting to force myself to pull Python out of my bag in place of Perl these days. Despite contributors to cpan doing what they can to fill the holes it is just an uphill battle in this world of cloud tools and apis everywhere when nobody releases a Perl version of anything.
Anecdotally I can assure you that javascript is EXTREMELY popular, it is so popular in fact that entire classes of job description simply omit it as an assumed requirement just like the ability to handle a few windows systems is assumed when hiring a Unix guy.
Of course, not to mention technical jobs. People are always so naive on these things. They obsess over incredibly expensive robots that can perform some sort of manual labor when it is NN 'bots' replacing high paid jobs they should worry about. Why spend all that money cutting out your lowest paid jobs when you can spend it cutting out some of your highest paid ones?
"At least in the US there's a history of these kind of projects"
There is a pretty well established history for both actually. I'm certainly not claiming the US are the good guys vs Russia who are the bad guys, Putin, China, North Korea are all countries with a long public record of being bad actors and there is little to no credible dispute of that. People do dispute the US, especially on a US forum like this one. As seen by someone replying and trying to paint the suggestion as a crackpot conspiracy theory.
Putin's crimes are just far more thinly veiled with a history of poorly covered domestic murder and death squads. The US on the other hand has aired much of its historical dirty laundry and so far has convinced the public of the "this is how we used to do things back in the wild west of history" dialog. Why people just blindly accept this is the past and anyone suggesting the US government has continued with business as usual is crazy and paranoid I don't know. Maybe it's just some very effective propaganda.
I think you are confusing the API, the abstracted standard and documentation which defines the interface, with the coded definition in their implementation. It's understandable because there will certainly be a massive amount of overlap in a new clean room implementation using that abstracted standard.
This is like a new state coming up with a drivers license and another state claiming it infringes on their copyright because it contains a picture, a dob, and a hologram and they look strikingly similar to that states and interoperability doesn't apply because the image in the hologram is different. In reality the new state took written descriptions of the security measures used by other states and came up with their own but in practice any implementation of those standards is going to end up looking similar. There are only so many ways to implement a clear head shot and so many ways to display a DOB. The same is true here but to an even greater extent, given there are common coding styles and practices a good portion of it likely would end up being identical character for character.
"The supreme court already was asked to rule and that point, and they let the appellate court's ruling stand. "
The story we are commenting under is a new request that isn't settled and most of the circuits currently agree that API's can't be copyrighted including the one that matters.
"and there is no legal theory to distinguish APIs from any other part of code"
APIs aren't part of the code. You really shouldn't comment on topics you are ignorant of and really neither should courts but sadly we are stuck with whatever random nonsense they spew from ignorance.