Yup. Everyone seems to get very excited about the pressure differential. It might be moderately difficult to maintain the low pressure in such a long pipe, but a failure would be very undramatic.
The land speed record is higher than 700 mph. When they finish their run they turn off the thrust and decelerate by friction with atmospheric pressure. In fact, they pop some chutes to increase the deceleration.
Have you ever dived down to 10 feet? The pressure difference on you in that situation is more than the difference a proposed hyperloop tube would experience.
The land speed record (in open air) is well in excess of 700 mph. Moving through 1 atmosphere at 700 mph isn't especially dramatic. It's just difficult and expensive to get something up to that speed and keep it going. You wouldn't have to design your trains to survive a leak. They just wouldn't be able to go as fast if there was one.
99% of plane crashes aren't catastrophic failures. The engines quitting and the plane gliding to a landing, or ditching in the ocean, are like a hyperloop train losing power and slowing down to an eventual stop.
A hyperloop train coming off its rails at full speed is like an airliner slamming into terrain at full speed. It's quite rare to survive that.
The ultrasound devices are fairly tight beams. The microwave ones probably are too, and it doesn't take much energy to shake up someone's brain, if you're depositing it effectively on the inside.
If you got a constructive node in a wall it might shake some dust loose. Your chances of hitting something fragile would be pretty small.
If you took that DARPA device that's designed to communicate over a kilometre and aimed it at someone's head from the other side of a hotel wall....
"But just to give you an intuition: Weak AI cannot plan, cannot judge, cannot explore"
Some machine learning control systems are explicitly designed to form and execute plans. The ones that are designed to be trained in the real world are usually also designed to learn to do internal simulations (imagining what will happen if I do X) because feedback in the real world is slow.
Judging is pretty much what machine learning systems do.
Many reinforcement learning methods have a hyperparameter to explicitly control the amount of exploration the system does.
Are you sure you're not thinking of 1960's era tech?
You've claimed it. Lots of people do. Now prove it.
Research has shown that our subjective assessment of our insight, use of logic, self-awareness and free will is a gross overestimate, at best. That's not to say we don't actually do those things, but there isn't really any proof that we do. And some of them are pretty problematic from what we know of physics.
Cent means 100 in latin. "Cent" in English properly refers to a unitless fraction, 1/100. You seem to assume that the word refers to 1/100th of a US dollar, which is kind of odd.
Will they make less? I think there's a good chance that the current advertising market is the result of runaway selection. Like a peacock. The advertisers might be better off if they were all forced to use simple, cheap, generic ads rather than all having to throw money at tracking, maintaining giant databases, data analysis research, all just so they don't fall slightly behind their competitors.
The US has lots of qualifications on freedom of speech codified into law. It starts in the US constitution ("Congress shall make no law....") and the infamous "the", and continues with lots of clarifying laws and court decisions.
You're correct that lots of other countries also have freedom of speech, with some restrictions, codified in their laws though.
It seems to me the American constitution says things about "inalienable" rights. They're not very inalienable if they only count when you're a US citizen on US soil, are they?
This is an irritating habit of machine learning researchers who aren't in much contact with the field they're applying algorithms to.
The most elementary error is to not balance your classes and report accuracy. As pointed out by the OP, this is meaningless and often extremely misleading.
Better is to balance your classes. This tells you about the performance of the algorithm, but only on a population with equal prevalence... i.e. not the real world.
Best is to give sensitivity and specificity and/or predictive values, or equivalent (ROC curve, likelihood ratios, etc.) on a sample that reflects real-world prevalence.
A classifier trained on a sample with 50/50 gay/straight split will almost certainly overestimate gay people in a natural dataset, compared to its performance in the unnatural one.
Identity theft seems to be an Americanism. I've rarely heard in mentioned in Canada. Of course, we also don't give out our equivalent of the SSN like candy. That number is government property and there are strong laws protecting who's allowed to have it.
I'm Canadian (although mostly Scottish by ancestry). I once stayed in a bed and breakfast in Ireland.
The woman running it said "oh, you're American!" I said "well, Canadian actually." She said "basically the same thing." I said "sure, I guess that makes you English?"
Bitcoin was also supposed to eliminate the big transaction fees charged by banks. But in order to get a transaction processed you have to pay fees that even the banks might blush at.
The scariest thing though is that lots and lots of bitcoin is held by a few random anonymous nerds. If one of those guys gets greedy, desperate or bored and tries to cash out....
Slashdot Mirror
Don't worry, if Amazon moved 50,000 employees to Montreal, they would all have to speak French (at work at least).
Also, they would have to change their name to Amazon: La Librarie.
Yup. Everyone seems to get very excited about the pressure differential. It might be moderately difficult to maintain the low pressure in such a long pipe, but a failure would be very undramatic.
Or you could use steel. Like a pipeline.
You know several companies have done the cost analysis right? I mean, actually done it, not just made up numbers on Slashdot.
The land speed record is higher than 700 mph. When they finish their run they turn off the thrust and decelerate by friction with atmospheric pressure. In fact, they pop some chutes to increase the deceleration.
The driver almost always survives.
Have you ever walked through one of those underwater acrylic tunnels they have in aquariums?
Have you ever dived down to 10 feet? The pressure difference on you in that situation is more than the difference a proposed hyperloop tube would experience.
Massive isn't really the word I'd use.
The land speed record (in open air) is well in excess of 700 mph. Moving through 1 atmosphere at 700 mph isn't especially dramatic. It's just difficult and expensive to get something up to that speed and keep it going. You wouldn't have to design your trains to survive a leak. They just wouldn't be able to go as fast if there was one.
99% of plane crashes aren't catastrophic failures. The engines quitting and the plane gliding to a landing, or ditching in the ocean, are like a hyperloop train losing power and slowing down to an eventual stop.
A hyperloop train coming off its rails at full speed is like an airliner slamming into terrain at full speed. It's quite rare to survive that.
The ultrasound devices are fairly tight beams. The microwave ones probably are too, and it doesn't take much energy to shake up someone's brain, if you're depositing it effectively on the inside.
If you got a constructive node in a wall it might shake some dust loose. Your chances of hitting something fragile would be pretty small.
If you took that DARPA device that's designed to communicate over a kilometre and aimed it at someone's head from the other side of a hotel wall....
"But just to give you an intuition: Weak AI cannot plan, cannot judge, cannot explore"
Some machine learning control systems are explicitly designed to form and execute plans. The ones that are designed to be trained in the real world are usually also designed to learn to do internal simulations (imagining what will happen if I do X) because feedback in the real world is slow.
Judging is pretty much what machine learning systems do.
Many reinforcement learning methods have a hyperparameter to explicitly control the amount of exploration the system does.
Are you sure you're not thinking of 1960's era tech?
You've claimed it. Lots of people do. Now prove it.
Research has shown that our subjective assessment of our insight, use of logic, self-awareness and free will is a gross overestimate, at best. That's not to say we don't actually do those things, but there isn't really any proof that we do. And some of them are pretty problematic from what we know of physics.
Unless of course it's an experiment and not a shitty observational study.
Cent means 100 in latin. "Cent" in English properly refers to a unitless fraction, 1/100. You seem to assume that the word refers to 1/100th of a US dollar, which is kind of odd.
There are a LOT of Safari browser users, and they have (and spend) more money on average than other browser users. iOS uses Safari.
Will they make less? I think there's a good chance that the current advertising market is the result of runaway selection. Like a peacock. The advertisers might be better off if they were all forced to use simple, cheap, generic ads rather than all having to throw money at tracking, maintaining giant databases, data analysis research, all just so they don't fall slightly behind their competitors.
Google used to. Google used to be famous for unobtrusive, text only ads.
Then they bought DoubleClick.
"Burning coal does not produce water."
Interesting! So where does the hydrogen go?
The US has lots of qualifications on freedom of speech codified into law. It starts in the US constitution ("Congress shall make no law....") and the infamous "the", and continues with lots of clarifying laws and court decisions.
You're correct that lots of other countries also have freedom of speech, with some restrictions, codified in their laws though.
It seems to me the American constitution says things about "inalienable" rights. They're not very inalienable if they only count when you're a US citizen on US soil, are they?
This is an irritating habit of machine learning researchers who aren't in much contact with the field they're applying algorithms to.
The most elementary error is to not balance your classes and report accuracy. As pointed out by the OP, this is meaningless and often extremely misleading.
Better is to balance your classes. This tells you about the performance of the algorithm, but only on a population with equal prevalence... i.e. not the real world.
Best is to give sensitivity and specificity and/or predictive values, or equivalent (ROC curve, likelihood ratios, etc.) on a sample that reflects real-world prevalence.
A classifier trained on a sample with 50/50 gay/straight split will almost certainly overestimate gay people in a natural dataset, compared to its performance in the unnatural one.
Identity theft seems to be an Americanism. I've rarely heard in mentioned in Canada. Of course, we also don't give out our equivalent of the SSN like candy. That number is government property and there are strong laws protecting who's allowed to have it.
Sounds like pharma in the 80s. Company offers some "samples" and maybe some kickbacks in return for good recommendations to the patients/students.
Today at least you have to keep it on the down low. In medicine.
I'm Canadian (although mostly Scottish by ancestry). I once stayed in a bed and breakfast in Ireland.
The woman running it said "oh, you're American!" I said "well, Canadian actually." She said "basically the same thing." I said "sure, I guess that makes you English?"
You don't have to know what you're doing. In fact, it's better if you don't "know what you're doing."
Buy index funds or throw darts at the financial pages. Don't ever check the value of your stocks.
Bitcoin was also supposed to eliminate the big transaction fees charged by banks. But in order to get a transaction processed you have to pay fees that even the banks might blush at.
The scariest thing though is that lots and lots of bitcoin is held by a few random anonymous nerds. If one of those guys gets greedy, desperate or bored and tries to cash out....