sharing of "intimate images" (defined) that were taken without consent is what is prohibited:
Not quite. Not only the "taking", but also the "distribution" must be consensual. So just because your GF lets you photo her boobs, that doesn't give you the right to share the photo with others.
the vast majority of incidents will never see any enforcement.
That is the way it should be. Unless there is an actual complaint, who cares? If there is a complaint, you can't weasel out of it by saying "She said it was okay".
Issuing orders that won't be followed helps destroy one's own authority.
Not really. Just like with civilian laws, there are some that are taken seriously, and others that are routinely ignored.
Disclaimer: I was a Marine, but that was back before photography had been invented. Semper Fi.
What we're really finding out here is that we need to build an orbital cleanup satellite.
We don't even need a satellite. We can do it from the earth's surface. Just put a big fricken laser on the summit of Mauna Kea or Cerro Toco, point it westward, and shoot it at the leading surface of the debris to slow it down into an unstable orbit.
There was no "relocation". The factory in Everett, Washington did not close, and Boeing had, and still has, no plans to close it. The issue was where a second factory would be built. Boeing offered to build it in Washington, if the union would agree to a long term contract with a no-strike clause. The union refused, and after a long legal fight, Boeing created the new jobs elsewhere. Some went to South Carolina, and some went overseas.
So you think it is okay for a company to close a plant in a state where workers have rights and moved to a state where workers can be abused with twice the hours at the same rate of pay?
Yes, I think that companies should be able to locate jobs in any state they choose. Overtime laws in Washington and South Carolina are similar, so I think you are spouting nonsense about that. Hourly pay for Boeing's assembly line workers in SC is about 20% lower than in WA, but the cost of living is also considerably lower.
The days when Boeing made most of the airplane in the US are long gone, and have been for a while.
That was because they tried to open a new factory in America, and the Obama administration said it was illegal for them to build a factory in a right-to-work state (which is most of them).
In other news, people who did bad things in the past continue to deny having done bad things.
The Turks don't deny that bad things happened, nor do they deny that they were the ones that did it.
What they do deny is that it amounted to genocide. In 1915 the British Empire landed 250,000 troops on the Gallipoli Peninsula, and the Russians launched a major offensive in the Caucasus. The Turks were fighting for their survival as a nation. The Gallipoli landing failed, mostly due to astoundingly incompetent leadership on the allied side, but also due to the brilliant and decisive leadership of Mustafa Kemal on the Turkish side. But the Russian offensive was successful, partly due to support from Armenians who viewed the Russians as liberators. In "liberated" areas, thousands of Armenians volunteered to fight with the Russians.
In the face of this offensive, and clear evidence that the Armenians were a "fifth column", there were Turkish reprisals and massacres, but mostly the Turks tried to deal with the Armenians by deporting them to Northern Syria (which was then part of their empire). This was all done during the deprivations of a world war, and many, many Armenians died enroute to Syria and in the camps once they got there.
Did bad things happen to the Armenians? Yes. Did really REALLY bad things happen to the Armenians? Yes. Did millions of Armenians die as a result of Turkish actions? Yes. Did this amount to a centrally planned and coordinated effort to exterminate the Armenian people? I don't think so.
Since Armenia is a landlocked country mostly surrounded by neighbors that hate them even more than the Turks do, and Turkey is their main route to the sea and one of their biggest trading partners, is it really in the best interest of the Armenian people to focus on "re-labelling" atrocities that happened over a century ago, rather than facing up to the many many problems that face Armenia today? You decide.
No. A force does not require energy. Only moving against a force requires energy. E=F*D. A newton is force, but a newton-meter is energy. So the magnet on your refrigerator does not use energy, but energy is required to pull it off.
How is it obtained, stored, replenished?
Here is a really cool fact that you can use to impress chicks at cocktail parties: A magnetic force and an electrical force are the SAME THING. The only difference is your inertial frame of reference. Let's say you have two parallel copper wires with current flowing through them. The negative charge in the electrons and the positive charge in the copper nuclei should cancel each other out, and there should be no force between them. BUT THERE IS. This is magnetism. But it is really just plain only electrical attraction because the electrons are moving, so their inertial reference frame is different from the reference frame of the copper nuclei. A moving reference frame has a Lorentz contraction, so the copper nuclei "see" more electrons per length of wire, resulting in an attraction.
TFA seems to leave out a lot of important geeky details. Like which supercomputer was used? How many hours of CPU (or maybe GPU?) time was used? Since they were running hundreds of models in parallel, why did they need a supercomputer at all? Wouldn't it have been more cost effective to rent compute servers in the cloud?
It is a poorly written summary, and not quite accurate
I don't think it is accurate at all. The guy spouting hype is the CEO of Textio, which is selling the very thing he is hyping. If any of this was actually happening at scale, productivity growth would be soaring. It is not. It is stagnating.
We made a simple center of gravity algorithm that calculates where a CPG company should place its next distribution center... This sort of AI is not hard if you know how to code
How is this "AI"? It sounds like you used a hand coded an algorithm, without any machine learning or any other AI techniques.
So matters are even worse than the articles suggest.
Not necessarily. The study is an example of "selection bias". They only looked at cases where the patient requested a 2nd opinion. So these were cases where the patient already had a reason to question the 1st opinion. It is unlikely that the error (or "different diagnosis") rate would be as high for other cases.
Before I go to a doctor, or take a family member, I do some online research to inform myself. If the doctor's diagnosis is significantly different from what I concluded, I ask for an explanation. If the doctor's "explanation" is that he is the expert and I should just trust him, then that usually means he is full of crap and doesn't know what he is doing.
Imagine the impact if the company was reviewed and then it was revealed that a fix for a security problem wasn't put in place due to the process.
Bureaucracies don't work that way. There are almost never negative consequences for inaction, especially if the proposal was never explicitly rejected. If someone comes to you with a proposal, you just bury it in some file drawer, or pass the decision (and blame) on to someone else.
When they say they're going to nuke us, and they're really trying to develop the needed technology, it might be worth taking them at their word.
They don't say they are going to nuke us. They say their nukes are for defensive purposes against American belligerence. Let's look at America's track record:
1. Saddam Hussein gave up his nuclear program. Result: America killed him.
2. Muammar Gaddafi gave up his nuclear program. Result: America killed him.
So how does it make any logical sense for Kim to give up his nukes? If America wants non-proliferation, maybe we should stop killing people who cooperate.
Given a choice, I would give up indoor plumbing before I would give up the Internet.... and yes, I have lived without indoor plumbing... for two years. I slept in my van, peed in the woods, pooped in an outhouse, and showered at work.
For the first 40 years, we didn't have fast enough hardware.
We also didn't have enough training data. Today, if you want to develop a NN to recognize faces, you can find petabytes of examples online. A decade ago, there was far less. 20 years ago, there was almost nothing.
Yes. Noise is inserted into the training data. Then the AI takes the data and extracts the relevant features by removing the noise. This is done by having one or more intermediate layers that are much smaller than the input layer. This forces the network to learn only the important features.
... also, even if all of that fails, because the stop sign has been defaced or even REMOVED COMPLETELY, the SDC will STOP ANYWAY. A database of the GPS coordinates of every stop sign in America will fit in $1 worth of flash. So even if the AI fails to recognize a stop sign, the software will know that a stop sign is supposed to be there and stop anyway.
But they can only learn if humans tech them a broad range of adulterated traffic signs.
Yup. The problem is called over-fitting, when the AI does well on the training data but not with real life data, and one way to mitigate that is "denoising", where random noise is injected into the training data. There are many other mitigation techniques such as dropout and ensemble bagging.
If your AI is confused by one sticker on a stop sign then you are not a competent developer. If there are a lot of stickers, then that may confuse a human as well... and, as has been pointed out many times before, self-driving cars don't have to be perfect, they just need to be better than humans.
This ad wasn't harmful, but it exposes what possibly could be done if someone wanted to be malicious.
So it is funny, harmless, and educational. That is even better.
There's a lot of malice that could be carried out if someone wanted to
Yes, people can do bad things. That doesn't mean that doing things is bad.
If anyone should be criticised here, it is Google, not BK. They should have some extra security, such as learning to recognize the voices of authorized users, or requiring an extra code word for purchases or IoT commands (basically anything other than just a request for info or to play a song).
Disclaimer: I have a Google Home and I am mostly happy with it.
sharing of "intimate images" (defined) that were taken without consent is what is prohibited:
Not quite. Not only the "taking", but also the "distribution" must be consensual. So just because your GF lets you photo her boobs, that doesn't give you the right to share the photo with others.
the vast majority of incidents will never see any enforcement.
That is the way it should be. Unless there is an actual complaint, who cares? If there is a complaint, you can't weasel out of it by saying "She said it was okay".
Issuing orders that won't be followed helps destroy one's own authority.
Not really. Just like with civilian laws, there are some that are taken seriously, and others that are routinely ignored.
Disclaimer: I was a Marine, but that was back before photography had been invented. Semper Fi.
New construction on Mauna Kea is already at a standstill because of some Hawaiians who think that telescopes disrespect the great volcano spirit
Nobody really believes that. They are just looking for a payoff. We just need to bribe the right people and the objections will disappear.
How is a laser in geostationary orbit cheaper than one on the ground ?
It converts solar radiation into electricity into laser light, without loss [because no atmosphere].
The summit of Cerro Toco is over 18,000 ft (5400 m) with near 0% humidity and 0% cloud cover. That solves 90% of the "atmosphere problem".
What we're really finding out here is that we need to build an orbital cleanup satellite.
We don't even need a satellite. We can do it from the earth's surface. Just put a big fricken laser on the summit of Mauna Kea or Cerro Toco, point it westward, and shoot it at the leading surface of the debris to slow it down into an unstable orbit.
No, the timing of the relocation ...
There was no "relocation". The factory in Everett, Washington did not close, and Boeing had, and still has, no plans to close it. The issue was where a second factory would be built. Boeing offered to build it in Washington, if the union would agree to a long term contract with a no-strike clause. The union refused, and after a long legal fight, Boeing created the new jobs elsewhere. Some went to South Carolina, and some went overseas.
So you think it is okay for a company to close a plant in a state where workers have rights and moved to a state where workers can be abused with twice the hours at the same rate of pay?
Yes, I think that companies should be able to locate jobs in any state they choose.
Overtime laws in Washington and South Carolina are similar, so I think you are spouting nonsense about that.
Hourly pay for Boeing's assembly line workers in SC is about 20% lower than in WA, but the cost of living is also considerably lower.
Which is why bars serve free peanuts, pretzels and chips - obviously they make their real money from food.
Beer has a lot of calories.
It is liquid bread.
The days when Boeing made most of the airplane in the US are long gone, and have been for a while.
That was because they tried to open a new factory in America, and the Obama administration said it was illegal for them to build a factory in a right-to-work state (which is most of them).
Liberal organizations are generally working for the common good.
Liberal organizations generally have good intentions.
That is not the same as working for the common good.
In other news, people who did bad things in the past continue to deny having done bad things.
The Turks don't deny that bad things happened, nor do they deny that they were the ones that did it.
What they do deny is that it amounted to genocide. In 1915 the British Empire landed 250,000 troops on the Gallipoli Peninsula, and the Russians launched a major offensive in the Caucasus. The Turks were fighting for their survival as a nation. The Gallipoli landing failed, mostly due to astoundingly incompetent leadership on the allied side, but also due to the brilliant and decisive leadership of Mustafa Kemal on the Turkish side. But the Russian offensive was successful, partly due to support from Armenians who viewed the Russians as liberators. In "liberated" areas, thousands of Armenians volunteered to fight with the Russians.
In the face of this offensive, and clear evidence that the Armenians were a "fifth column", there were Turkish reprisals and massacres, but mostly the Turks tried to deal with the Armenians by deporting them to Northern Syria (which was then part of their empire). This was all done during the deprivations of a world war, and many, many Armenians died enroute to Syria and in the camps once they got there.
Did bad things happen to the Armenians?
Yes.
Did really REALLY bad things happen to the Armenians?
Yes.
Did millions of Armenians die as a result of Turkish actions?
Yes.
Did this amount to a centrally planned and coordinated effort to exterminate the Armenian people?
I don't think so.
Since Armenia is a landlocked country mostly surrounded by neighbors that hate them even more than the Turks do, and Turkey is their main route to the sea and one of their biggest trading partners, is it really in the best interest of the Armenian people to focus on "re-labelling" atrocities that happened over a century ago, rather than facing up to the many many problems that face Armenia today?
You decide.
Exerting force requires energy, no?
No. A force does not require energy. Only moving against a force requires energy. E=F*D. A newton is force, but a newton-meter is energy. So the magnet on your refrigerator does not use energy, but energy is required to pull it off.
How is it obtained, stored, replenished?
Here is a really cool fact that you can use to impress chicks at cocktail parties: A magnetic force and an electrical force are the SAME THING. The only difference is your inertial frame of reference. Let's say you have two parallel copper wires with current flowing through them. The negative charge in the electrons and the positive charge in the copper nuclei should cancel each other out, and there should be no force between them. BUT THERE IS. This is magnetism. But it is really just plain only electrical attraction because the electrons are moving, so their inertial reference frame is different from the reference frame of the copper nuclei. A moving reference frame has a Lorentz contraction, so the copper nuclei "see" more electrons per length of wire, resulting in an attraction.
TFA seems to leave out a lot of important geeky details. Like which supercomputer was used? How many hours of CPU (or maybe GPU?) time was used? Since they were running hundreds of models in parallel, why did they need a supercomputer at all? Wouldn't it have been more cost effective to rent compute servers in the cloud?
It is a poorly written summary, and not quite accurate
I don't think it is accurate at all. The guy spouting hype is the CEO of Textio, which is selling the very thing he is hyping. If any of this was actually happening at scale, productivity growth would be soaring. It is not. It is stagnating.
We made a simple center of gravity algorithm that calculates where a CPG company should place its next distribution center ... This sort of AI is not hard if you know how to code
How is this "AI"? It sounds like you used a hand coded an algorithm, without any machine learning or any other AI techniques.
So matters are even worse than the articles suggest.
Not necessarily. The study is an example of "selection bias". They only looked at cases where the patient requested a 2nd opinion. So these were cases where the patient already had a reason to question the 1st opinion. It is unlikely that the error (or "different diagnosis") rate would be as high for other cases.
Before I go to a doctor, or take a family member, I do some online research to inform myself. If the doctor's diagnosis is significantly different from what I concluded, I ask for an explanation. If the doctor's "explanation" is that he is the expert and I should just trust him, then that usually means he is full of crap and doesn't know what he is doing.
Imagine the impact if the company was reviewed and then it was revealed that a fix for a security problem wasn't put in place due to the process.
Bureaucracies don't work that way. There are almost never negative consequences for inaction, especially if the proposal was never explicitly rejected. If someone comes to you with a proposal, you just bury it in some file drawer, or pass the decision (and blame) on to someone else.
Odds are any company with young people running it will be gone in a decade. Not worth anyone's time.
Like Facebook (Mark was 19), Google (Larry was 25), and Microsoft (Bill was 20)?
When they say they're going to nuke us, and they're really trying to develop the needed technology, it might be worth taking them at their word.
They don't say they are going to nuke us. They say their nukes are for defensive purposes against American belligerence. Let's look at America's track record:
1. Saddam Hussein gave up his nuclear program.
Result: America killed him.
2. Muammar Gaddafi gave up his nuclear program.
Result: America killed him.
So how does it make any logical sense for Kim to give up his nukes?
If America wants non-proliferation, maybe we should stop killing people who cooperate.
Given a choice, I would give up indoor plumbing before I would give up the Internet. ... and yes, I have lived without indoor plumbing ... for two years.
I slept in my van, peed in the woods, pooped in an outhouse, and showered at work.
For the first 40 years, we didn't have fast enough hardware.
We also didn't have enough training data. Today, if you want to develop a NN to recognize faces, you can find petabytes of examples online. A decade ago, there was far less. 20 years ago, there was almost nothing.
But doesn't the "de-" prefix mean "remove from"?
Yes. Noise is inserted into the training data. Then the AI takes the data and extracts the relevant features by removing the noise. This is done by having one or more intermediate layers that are much smaller than the input layer. This forces the network to learn only the important features.
... also, even if all of that fails, because the stop sign has been defaced or even REMOVED COMPLETELY, the SDC will STOP ANYWAY. A database of the GPS coordinates of every stop sign in America will fit in $1 worth of flash. So even if the AI fails to recognize a stop sign, the software will know that a stop sign is supposed to be there and stop anyway.
But they can only learn if humans tech them a broad range of adulterated traffic signs.
Yup. The problem is called over-fitting, when the AI does well on the training data but not with real life data, and one way to mitigate that is "denoising", where random noise is injected into the training data. There are many other mitigation techniques such as dropout and ensemble bagging.
If your AI is confused by one sticker on a stop sign then you are not a competent developer. If there are a lot of stickers, then that may confuse a human as well ... and, as has been pointed out many times before, self-driving cars don't have to be perfect, they just need to be better than humans.
The bigger fish they should be frying should be the "crippling construction defects"
I may be going way out on a limb here, but maybe ship construction and health issues are handled by different people.
This ad wasn't harmful, but it exposes what possibly could be done if someone wanted to be malicious.
So it is funny, harmless, and educational. That is even better.
There's a lot of malice that could be carried out if someone wanted to
Yes, people can do bad things. That doesn't mean that doing things is bad.
If anyone should be criticised here, it is Google, not BK. They should have some extra security, such as learning to recognize the voices of authorized users, or requiring an extra code word for purchases or IoT commands (basically anything other than just a request for info or to play a song).
Disclaimer: I have a Google Home and I am mostly happy with it.