It's entirely possible to have good credit while not being in debt
Uh....no. To get a high FICO score, you're going to have to be in debt to somebody.
Just having a credit card that you pay off every month doesn't help, because a component of your FICO score is the mix of your credit accounts. If you don't have something like a mortgage or car loan, it is going to hurt your score.
Ah, I see: you're replying to the wrong thread. You're looking for the post by strawman. Were you replying to my thread, you'd have noticed I claimed the opposite.
Nope, you just cut off your discussion at Asians. Which, btw, is wrong statistically - more whites have the relevant jobs. The fact that you did not explicitly state your bell curves applied to women vs men didn't suddenly alter your overall thesis.
I see you've never worked with "big data". You can't make such firm claims about what's not in a very large data set
You really think it's impossible to take a data set that is 30% women and not be able to prove it's 30% women?
Or one that's 100% red cars and prove it's 0% blue cars?
Negative, you're asserting sexism. The case is yours to make.
The current situation is more women are employed in these roles than men. The simplest explanation is the people hiring and working to retain the workers are hiring and retaining more men than women. AKA sexism.
You're simply saying that the company is presumed guilty until proven innocent.
I'm saying you need to come up with a way to get the same results. Not just assert that there is no sexism.
"There's a dead body here. It's been shot three times." "Well, it couldn't possibly be murder".
Look, I'm okay with us being a society where guilt is presumed.
Just understand that I'm going to accuse you of pedophilia and you're going to be put on a sex offender list.
Ya know, there are a lot shorter ways to say "I can't come up with anything to explain why it isn't sexism".
So, push that button. We're already seeing this with the MeToo movement.
Oh, I see the problem. You view sexism as a feature.
That is illogical. There are many variables that could correlate with gender that are legitimate reasons for job selection. [...]
Or have we reversed burden of proof?
The null hypothesis would be that men and women are identical when it comes to working in non-strength-dependent jobs.
You are asserting that this isn't true because "variables" that you've failed to state.
The burden of proof is on you to show that men and women are not identical when it comes to working in non-strength-dependent jobs. As in state what those variables are and measure them such that you can prove your hypothesis.
And if you're going to point at the outcome (such as less women working in this field), then you need to go look up causation vs. correlation.
The point is, it doesn't matter how you establish it.
Uh...yeah, actually it does. Because looking at the results of a process don't tell you the cause. That whole "correlation vs causation" thing.
You are asserting that women and non-whites have their "bell curve" such that they're not as good as white men. Your evidence is the correlation that they are underrepresented in the workforce.
You've yet to do anything looking at causation for that difference.
You cannot, from the output of a machine learning system, prove that it isn't using X as criteria in screening candidates.
Sure. Feed it nothing but red cars, and it will never pick a blue car. Because you limited the input. Really easy to prove in court too.
Yes, you have to establish that such bell curves differ based on gender and not-whiteness. And again, if you're just looking at outcome as your proof, you're falling for the same incomplete data as the AI in TFS.
It's not about "incomplete data as the AI" unless the coders were complete idiots. Amazon has a huge population of developers.
That huge population was created with bias in mind, not only in initial hiring but in how women are treated in the workforce over the years. Thus you can not trust that this data is completely impartial.
It's the nature of machine learning systems that you can't really prove what criteria they've "learned" from your training data
Sure you can. The machine learning system will produce a result that closely matches the training data. When your training data is the result of bias, the machine learning system will strive to continue that bias.
If your training data only has red cars in it, your machine learning system isn't going to believe cars can be blue.
It is a lot easier to use baseload overcapacity to do things like pumped hydro to smooth daily demand fluctuations than it is to try and match an unpredictable supply to fluctuating and not completely predictable demand.
The flaw in this is there aren't that many places where you can do pumped hydro easily. You need a mountain with two lakes on it or similar terrain where your upper reservoir would run through a narrow channel. There's not many places where you have close to that structure naturally that aren't already being used for hydro.
So, you either spend a fortune building it, or you spend that same fortune buying a giant battery facility with essentially the same capacity.
Now think of that single person, planting something that accepts remote (like via satellite phone, or even cellular if in port) can now remotely own these billion dollar weapon platforms.
Now think about how they could do this even if you apply any security measures you can come up with.
"We kept him from hacking the phalanx system! Instead, he planted something that broke the engines, so the phalanx system is down because we have no power."
Purge any submission to the system of a gender identifier... women's or men's anything... remove names in case that is factored... literally provide nothing in the submission that would definitively define a gender.
Writing style can tend to be different between genders. And you can't really remove that without feeding it blank pages.
Pretending computers have a bias vs anything is really dumb
No, what's dumb is pretending the computer is doing anything beyond its training data.
You give the computer biased training data, you get biased results. Not because the computer has bias, but because your training data is the result of bias.
You can't add security on as an afterthought. It needs to be a core feature.
Adding security in and of itself is dangerous. If the operator can't fire the weapon because he's locked out of the terminal, it is worse than not having that weapon there at all. Because you make your plans assuming the weapon is present, and when it won't work then your plans are fucked.
Military security comes from people walking around with guns and not plugging everything into the Internet.
if the customer doesn't properly write in cyber as a requirement
There's a more important question: Is it proper?
Computerized gun turret. It is only connected to a network that goes to a small number of secure terminals, which are not connected to any other network.
Why do you need to encrypt that link? If you control physical access via people with guns, why do you need secure logins?
"We made this guy change his 16-character password every 2 months and he forgot it while getting shot at. Now he's locked out of the terminal due to three failed login attempts. The local sysadmin is dead, so he can't unlock it. It would be nice to return fire, but we gotta have network security on this air-gapped network!"
Military requirements are not the same as civilian requirements. There's a reason tanks do not have ignition keys like cars do.
That's really the dumb part of this story. These systems are air-gapped.
At that point, you have to decide if the air gap is enough or if you want to add more security. When making that decision, you have to consider things like "If we can't fire this when we need to because a certificate expired, we will die".
And "an operator could sabotage this" doesn't require hacking the computer. As you say, throw a wrench in it. Or unplug it. Or fill the operator's station with bullets.
Why would you put your trojan chip in the ethernet connector? It's away from the signals you want to get to - serial and/or usb wires to get at UEFI
Option 1: There's no requirement that there be no connection from this chip to another part of the computer. If you're planting rogue hardware in one thing there's little reason to believe you can't plant rogue hardware elsewhere.
Option 2: PXE boot attack.
Option 3: Their desired attack vector only uses this additional chip, the chip itself isn't the attack vector. After all, where do you get that BIOS image? Or the firmware on the rest of the motherboard's chips? Or perhaps you need the server to receive a malformed packet to implement an exploit of some other process, but that malformed packet can't survive routing.
Option 4: This isn't for spying. Cut off all packets or scramble some packets and you've effectively disabled the server.
It's not like insurance companies don't already have your health information on file, now is it?
If your employer self-insures (virtually all that have >200 employees do), then your employer has your health information too.
Subject to various state regulations, and there are supposed to be restrictions on who can see that data and what they can do with it. Heck, they may only get aggregate information for the entire company. But if a pricey bill for chemotherapy comes in and Jim starts taking a lot of vacation time on those treatment dates, not really hard to put two and two together.
Clearly if Uber were purchased by Microsoft, the drivers would no longer be a core part of the company's business
Companies can have multiple lines of business. For example, Microsoft sells hardware, software and cloud services. It doesn't matter if one of those groups makes the most money, Microsoft's "core business" include all three.
Is there any cloud service in the world that is secure against hackers/malware?
Yes.
Step 1: Don't connect it to the regular Internet. And before you think that's unworkable, the DoD currently runs three "Internet" networks. NIPRNet, SIPRNet and JWICS.
Slashdot Mirror
We graduate 1.5 STEM students from college for every entry-level STEM job opening.
Now, tell me again how going to college and picking the supposedly "hot" degree will ensure high pay.
It's entirely possible to have good credit while not being in debt
Uh....no. To get a high FICO score, you're going to have to be in debt to somebody.
Just having a credit card that you pay off every month doesn't help, because a component of your FICO score is the mix of your credit accounts. If you don't have something like a mortgage or car loan, it is going to hurt your score.
Ah, I see: you're replying to the wrong thread. You're looking for the post by strawman. Were you replying to my thread, you'd have noticed I claimed the opposite.
Nope, you just cut off your discussion at Asians. Which, btw, is wrong statistically - more whites have the relevant jobs. The fact that you did not explicitly state your bell curves applied to women vs men didn't suddenly alter your overall thesis.
I see you've never worked with "big data". You can't make such firm claims about what's not in a very large data set
You really think it's impossible to take a data set that is 30% women and not be able to prove it's 30% women?
Or one that's 100% red cars and prove it's 0% blue cars?
Negative, you're asserting sexism. The case is yours to make.
The current situation is more women are employed in these roles than men. The simplest explanation is the people hiring and working to retain the workers are hiring and retaining more men than women. AKA sexism.
You're simply saying that the company is presumed guilty until proven innocent.
I'm saying you need to come up with a way to get the same results. Not just assert that there is no sexism.
"There's a dead body here. It's been shot three times." "Well, it couldn't possibly be murder".
Look, I'm okay with us being a society where guilt is presumed.
Just understand that I'm going to accuse you of pedophilia and you're going to be put on a sex offender list.
Ya know, there are a lot shorter ways to say "I can't come up with anything to explain why it isn't sexism".
So, push that button. We're already seeing this with the MeToo movement.
Oh, I see the problem. You view sexism as a feature.
That is illogical. There are many variables that could correlate with gender that are legitimate reasons for job selection. [...]
Or have we reversed burden of proof?
The null hypothesis would be that men and women are identical when it comes to working in non-strength-dependent jobs.
You are asserting that this isn't true because "variables" that you've failed to state.
The burden of proof is on you to show that men and women are not identical when it comes to working in non-strength-dependent jobs. As in state what those variables are and measure them such that you can prove your hypothesis.
And if you're going to point at the outcome (such as less women working in this field), then you need to go look up causation vs. correlation.
The point is, it doesn't matter how you establish it.
Uh...yeah, actually it does. Because looking at the results of a process don't tell you the cause. That whole "correlation vs causation" thing.
You are asserting that women and non-whites have their "bell curve" such that they're not as good as white men. Your evidence is the correlation that they are underrepresented in the workforce.
You've yet to do anything looking at causation for that difference.
You cannot, from the output of a machine learning system, prove that it isn't using X as criteria in screening candidates.
Sure. Feed it nothing but red cars, and it will never pick a blue car. Because you limited the input. Really easy to prove in court too.
It's not all-or-nothing, it's bell curves.
Yes, you have to establish that such bell curves differ based on gender and not-whiteness. And again, if you're just looking at outcome as your proof, you're falling for the same incomplete data as the AI in TFS.
It's not about "incomplete data as the AI" unless the coders were complete idiots. Amazon has a huge population of developers.
That huge population was created with bias in mind, not only in initial hiring but in how women are treated in the workforce over the years. Thus you can not trust that this data is completely impartial.
It's the nature of machine learning systems that you can't really prove what criteria they've "learned" from your training data
Sure you can. The machine learning system will produce a result that closely matches the training data. When your training data is the result of bias, the machine learning system will strive to continue that bias.
If your training data only has red cars in it, your machine learning system isn't going to believe cars can be blue.
The current Republican "party", which bears little resemblance to the real Republican party of years gone by, has been lying to their base for years.
It's been many decades. It's not that they're failing to live up to an ideal. This is who they are.
It is a lot easier to use baseload overcapacity to do things like pumped hydro to smooth daily demand fluctuations than it is to try and match an unpredictable supply to fluctuating and not completely predictable demand.
The flaw in this is there aren't that many places where you can do pumped hydro easily. You need a mountain with two lakes on it or similar terrain where your upper reservoir would run through a narrow channel. There's not many places where you have close to that structure naturally that aren't already being used for hydro.
So, you either spend a fortune building it, or you spend that same fortune buying a giant battery facility with essentially the same capacity.
If the safety disables the gun until a tech unlocked it, yes we should not have those.
If the safety is a simple lever that any half-sentient being can successfully operate under extreme stress, then have the safety.
Now think of that single person, planting something that accepts remote (like via satellite phone, or even cellular if in port) can now remotely own these billion dollar weapon platforms.
Now think about how they could do this even if you apply any security measures you can come up with.
"We kept him from hacking the phalanx system! Instead, he planted something that broke the engines, so the phalanx system is down because we have no power."
Purge any submission to the system of a gender identifier... women's or men's anything... remove names in case that is factored... literally provide nothing in the submission that would definitively define a gender.
Writing style can tend to be different between genders. And you can't really remove that without feeding it blank pages.
Pretending computers have a bias vs anything is really dumb
No, what's dumb is pretending the computer is doing anything beyond its training data.
You give the computer biased training data, you get biased results. Not because the computer has bias, but because your training data is the result of bias.
In other words, if your only criteria is hiring whomever best for the job, you will likely be operating illegally and subject to fines and lawsuits.
Until you can demonstrate that women or non-whites are not physically capable of being "the best for the job", then this reasoning is flawed.
And if your "proof" is the lack of not-white-men in these positions, you're falling for the same incomplete data as the AI in TFS.
You can't add security on as an afterthought. It needs to be a core feature.
Adding security in and of itself is dangerous. If the operator can't fire the weapon because he's locked out of the terminal, it is worse than not having that weapon there at all. Because you make your plans assuming the weapon is present, and when it won't work then your plans are fucked.
Military security comes from people walking around with guns and not plugging everything into the Internet.
if the customer doesn't properly write in cyber as a requirement
There's a more important question: Is it proper?
Computerized gun turret. It is only connected to a network that goes to a small number of secure terminals, which are not connected to any other network.
Why do you need to encrypt that link? If you control physical access via people with guns, why do you need secure logins?
"We made this guy change his 16-character password every 2 months and he forgot it while getting shot at. Now he's locked out of the terminal due to three failed login attempts. The local sysadmin is dead, so he can't unlock it. It would be nice to return fire, but we gotta have network security on this air-gapped network!"
Military requirements are not the same as civilian requirements. There's a reason tanks do not have ignition keys like cars do.
That's really the dumb part of this story. These systems are air-gapped.
At that point, you have to decide if the air gap is enough or if you want to add more security. When making that decision, you have to consider things like "If we can't fire this when we need to because a certificate expired, we will die".
And "an operator could sabotage this" doesn't require hacking the computer. As you say, throw a wrench in it. Or unplug it. Or fill the operator's station with bullets.
Why would you put your trojan chip in the ethernet connector? It's away from the signals you want to get to - serial and/or usb wires to get at UEFI
Option 1: There's no requirement that there be no connection from this chip to another part of the computer. If you're planting rogue hardware in one thing there's little reason to believe you can't plant rogue hardware elsewhere.
Option 2: PXE boot attack.
Option 3: Their desired attack vector only uses this additional chip, the chip itself isn't the attack vector. After all, where do you get that BIOS image? Or the firmware on the rest of the motherboard's chips? Or perhaps you need the server to receive a malformed packet to implement an exploit of some other process, but that malformed packet can't survive routing.
Option 4: This isn't for spying. Cut off all packets or scramble some packets and you've effectively disabled the server.
Perhaps these devices were installed at the behest of the NSA and the Chinese simply redesigned them to also send info to the Chinese government.
Because when I want to implement my super-secret and highly illegal surveillance program, I turn to a hostile government to implement it. :eyeroll:
It's not like insurance companies don't already have your health information on file, now is it?
If your employer self-insures (virtually all that have >200 employees do), then your employer has your health information too.
Subject to various state regulations, and there are supposed to be restrictions on who can see that data and what they can do with it. Heck, they may only get aggregate information for the entire company. But if a pricey bill for chemotherapy comes in and Jim starts taking a lot of vacation time on those treatment dates, not really hard to put two and two together.
Clearly if Uber were purchased by Microsoft, the drivers would no longer be a core part of the company's business
Companies can have multiple lines of business. For example, Microsoft sells hardware, software and cloud services. It doesn't matter if one of those groups makes the most money, Microsoft's "core business" include all three.
because an Uber driver is under no obligation to accept any available drive.
Except they are. Don't accept some rides and you get kicked off the app.
Doesn't matter what Uber says, it matters what Uber does.
Employers never let their employees work for a competitor while still working for them.
This is false both for contractors and employees.
You can have a job at McDonalds and a second job at Wendy's.
Also, a contractor can agree to not do any work for a specified list of companies when signing the contract with a company.
They are paid a portion of the fare, not a fixed hourly rate
And this lawsuit is over whether or not this practice is legal.
That is one of the IRS criteria for contractors vs employees.
Since this is a state lawsuit the IRS's criteria are irrelevant. Only CA labor law matters.
Is there any cloud service in the world that is secure against hackers/malware?
Yes.
Step 1: Don't connect it to the regular Internet. And before you think that's unworkable, the DoD currently runs three "Internet" networks. NIPRNet, SIPRNet and JWICS.