Well, I can totally see Google, Amazon or Facebook creating a new data center in the center of the earth and drilling a hole down from the North Pole for a cooling line.
OK, the Hudson Splash. But it would have been a crash if computers had been in control. Possibly into a populated area, too. Standard engine failure procedure is climb, usually straight ahead or along some trajectory that avoids obstacles, until reaching a safe altitude to accelerate, retract the flaps, and prepare for a new approach. An autopilot programmed to follow that procedure would simply keep going straight ahead (or along the programmed trajectory) while descending until it hit the ground. It would never even consider turning towards the water.
A quick Google search, however, tells me the US Air Force had 38 drone crashes in 2017. That was more than I even expected. And there aren't tens of thousands of them flying around non stop. So I think it's fair to say they are a lot less reliable than piloted airplanes.
Do you have any idea how many drones crash? If airliners had the same rate, nobody would dare to fly anymore. Yet these drones have much easier requirements. They fly a fixed route (possibly updated by remote control), outside of busy airspaces, and only in good weather. No complicated electrical and air conditioning systems to support passengers. Not even a need for flaps thanks to their low weight. If there's one kind of air transport that we ought to be able to automate, it's drones. But they crash all the time.
Yeah, I'd like to see an automated system handle a situation like Qantas 32. Or the Hudson crash.
And before someone goes "but most crashes are caused by pilot error": the vast majority of would-be crashes that would have been caused by automation are actually prevented by the pilots. Automation screws up all the time. In fact, many crashes that were caused by automation problems are actually classified as "pilot error" because the pilots should have been paying attention and prevented the crash. For example the Turkish Airlines crash in Amsterdam where the airplane stalled during a fully automatic approach, yet the pilots were blamed for not intervening when the airspeed dropped below approach speed. I have actually had a similar situation but reacted correctly, resulting in... an air safety report filed after landing. Didn't make the papers;-)
Another interesting bit was the maneuvers it was doing during the last seconds, tilting heavily before righting itself again. Looks like it was canceling a significant amount of horizontal velocity. Did it intentionally try to get as close to land as possible and cancel the horizontal speed at the last possible moment? Or was it just trying to land wherever it ended up? Certainly impressive how the control systems didn't give up and made the best of a bad situation. And especially with the rocket still spinning, being able to get the right tilt angle is amazing.
Not just pulling it out of the water, but more importantly cleaning out all the salt water and replacing the parts that got damaged by it. Salt water is very bad for rocket components.
1. Hidden subsidies for anything oil-related 2. Hidden price of pollution 3. Lack of economics of scale.
By using subsidies, you can increase volume so you solve number 3 and get 1 and 2 as a bonus.
Seriously, if you look at the complexity of a modern ICE engine, how can it possibly be cheaper? Only because the technology for producing enormous quantities of them has already been paid for.
Want to talk economics? China has lots of companies producing electric cars. The first versions were shit, but they are improving rapidly. Companies like BYD are building new factories all over the place. There's Chinese electric scooters (lots of them), Chinese electric sedans, Chinese electric sports cars, Chinese electric buses, Chinese electric garbage trucks, etcetera. Their technology is improving rapidly and volume is increasing exponentially. They are still nowhere near Tesla quality, but they'll get there eventually.
Meanwhile, what's the US doing? Hitting the brakes to preserve profits for old fashioned ICE engine manufacturers. Until one day, in a decade or so, they wake up to drastically better batteries so nobody wants a loud smelly ICE anymore. And you'll all be driving Chinese cars because the US is way behind the curve.
Way to go for short term economic benefit.
Same goes for green electricity. It's already becoming cheaper than coal, and one day it won't make sense anymore to keep burning stuff to generate electricity. Guess who will have the required know-how for these new technologies? Countries encouraging them, or countries trying to protect their vested interests in old technology? Good luck to the US.
Yeah, while other countries are still enforcing their naive ethics-based bans, the Chinese just power ahead. I, for one, welcome our new Chinese mutant overlords.
I wonder how much of this $7.8 billion figure is from bots gobbling up stocks of items to later resell at inflated prices. Apperently, according to another Slashdot article today, this is becoming a real problem and may well inflate the sales figure quite a bit.
I wonder what's best for the owner, though. Getting your car back in one piece without the thief, or getting it back with a thief inside who smashed the interior and windows trying to get out.
I don't know APK, and whether or not his complaints are valid, but his post might actuallty be unintentionally on topic. If reputation is a factor in this Chinese points program, and a few people with a grudge against you start "downmodding" you in this system, that can have pretty serious consequences. There was an article about a kid not being admitted at a school because his father's score was too low, for example.
Inventing is not just about having a concept like "maybe we can let a rocket land on its tail". It's also about actually making it work. Even the latest rocket being designed by NASA today, the SLS, is an expensive throwaway rocket. So what's the reason today?
Your advice to innovators seems to be "stop innovating, because you're just pretend engineers that are recycling old ideas". Fortunately, the actual innovators aren't listening to you and are actually pushing human society forward.
SpaceX is not on the stock market. And Elon has contemplated taking Tesla off the stock market because apparently he felt it was holding them back by always forcing them to look at the short term.
As for recycling 100 year old technology: why haven't Boeing, Lockheed and the others done reusable rockets then, instead of wasting billions of dollars making expensive non-reusable rockets that are four times as expensive as even a non-reused Falcon 9? After all, it's not like this is brand new and exciting technology, they could have done so ages ago. With all the money NASA would have saved, they could have been on Mars by now. So why haven't they?
Or maybe the technology wasn't quite so simple and they all thought it was impossible.
Yes, my classical computer is perfectly capable of processing more than 2^17,000,000,000 different states just counting its RAM. I don't see what the problem is. Hell, I can easily draw one of many googols of possible states on a sheet of paper. Doesn't keep me from processing it.
I understand quantum computing is a lot harder than classcial, but if difficulty scaled with the number of states, the 72 qubit quantum computer made by Google, with almost 5 sextillion (10^21) states, ought to already be impossible. Seems to be working just fine regardless.
This should be modded way up, even it it's an anonymous coward. The summary is indeed completely wrong, the radar actually works by comparing reflected signals to the signal that stayed home, thereby canceling background and jamming noise.
Doesn't matter how you do it. Entanglement cannot be used to send information. It can be used for encryption, though. Send entangled photons to two parties, they measure the photons so they each get the same, totally random key, and then they use that key to encrypt and decrypt a message sent via traditional means.
But doing something to the photons at one end and somehow directly affecting the other end is not possible. So you cannot tell whether or not a radar beam hit an aircraft by observing the entangled particles on the ground.
That's not how entanglement works. Maybe they are using some other kind of quantum effect. But entanglement actually does work instantaneously, faster than light. It just works in such a way that you cannot use it for communication. If you measure the particles at both ends, you will get the same result instantaneously (while you can prove that the decoherence occurred at that exact moment, not earlier, and therefore some kind of "information" must have traveled FTL) but you have no control over that result, and you cannot tell whether or not the other party made a measurement. So you cannot use it to send a message.
If this worked, it would allow faster than light communication. Send one half of an entangled pair of beams to someone (traveling at the speed of light, but that’s just the setup, not the actual communication), then the recipient disturbs the beam (or not), and the other party instantaneously notices. Nope, not possible.
Slashdot Mirror
Looks like they have a death wish if they mocked Erik Apple.
Well, I can totally see Google, Amazon or Facebook creating a new data center in the center of the earth and drilling a hole down from the North Pole for a cooling line.
OK, the Hudson Splash. But it would have been a crash if computers had been in control. Possibly into a populated area, too. Standard engine failure procedure is climb, usually straight ahead or along some trajectory that avoids obstacles, until reaching a safe altitude to accelerate, retract the flaps, and prepare for a new approach. An autopilot programmed to follow that procedure would simply keep going straight ahead (or along the programmed trajectory) while descending until it hit the ground. It would never even consider turning towards the water.
Fair point :-)
A quick Google search, however, tells me the US Air Force had 38 drone crashes in 2017. That was more than I even expected. And there aren't tens of thousands of them flying around non stop. So I think it's fair to say they are a lot less reliable than piloted airplanes.
Do you have any idea how many drones crash? If airliners had the same rate, nobody would dare to fly anymore. Yet these drones have much easier requirements. They fly a fixed route (possibly updated by remote control), outside of busy airspaces, and only in good weather. No complicated electrical and air conditioning systems to support passengers. Not even a need for flaps thanks to their low weight. If there's one kind of air transport that we ought to be able to automate, it's drones. But they crash all the time.
Yeah, I'd like to see an automated system handle a situation like Qantas 32. Or the Hudson crash.
And before someone goes "but most crashes are caused by pilot error": the vast majority of would-be crashes that would have been caused by automation are actually prevented by the pilots. Automation screws up all the time. In fact, many crashes that were caused by automation problems are actually classified as "pilot error" because the pilots should have been paying attention and prevented the crash. For example the Turkish Airlines crash in Amsterdam where the airplane stalled during a fully automatic approach, yet the pilots were blamed for not intervening when the airspeed dropped below approach speed. I have actually had a similar situation but reacted correctly, resulting in... an air safety report filed after landing. Didn't make the papers ;-)
Another interesting bit was the maneuvers it was doing during the last seconds, tilting heavily before righting itself again. Looks like it was canceling a significant amount of horizontal velocity. Did it intentionally try to get as close to land as possible and cancel the horizontal speed at the last possible moment? Or was it just trying to land wherever it ended up? Certainly impressive how the control systems didn't give up and made the best of a bad situation. And especially with the rocket still spinning, being able to get the right tilt angle is amazing.
Not just pulling it out of the water, but more importantly cleaning out all the salt water and replacing the parts that got damaged by it. Salt water is very bad for rocket components.
Until you open the box.
They're not cheaper yet, because of
1. Hidden subsidies for anything oil-related
2. Hidden price of pollution
3. Lack of economics of scale.
By using subsidies, you can increase volume so you solve number 3 and get 1 and 2 as a bonus.
Seriously, if you look at the complexity of a modern ICE engine, how can it possibly be cheaper? Only because the technology for producing enormous quantities of them has already been paid for.
Want to talk economics? China has lots of companies producing electric cars. The first versions were shit, but they are improving rapidly. Companies like BYD are building new factories all over the place. There's Chinese electric scooters (lots of them), Chinese electric sedans, Chinese electric sports cars, Chinese electric buses, Chinese electric garbage trucks, etcetera. Their technology is improving rapidly and volume is increasing exponentially. They are still nowhere near Tesla quality, but they'll get there eventually.
Meanwhile, what's the US doing? Hitting the brakes to preserve profits for old fashioned ICE engine manufacturers. Until one day, in a decade or so, they wake up to drastically better batteries so nobody wants a loud smelly ICE anymore. And you'll all be driving Chinese cars because the US is way behind the curve.
Way to go for short term economic benefit.
Same goes for green electricity. It's already becoming cheaper than coal, and one day it won't make sense anymore to keep burning stuff to generate electricity. Guess who will have the required know-how for these new technologies? Countries encouraging them, or countries trying to protect their vested interests in old technology? Good luck to the US.
Yeah, while other countries are still enforcing their naive ethics-based bans, the Chinese just power ahead. I, for one, welcome our new Chinese mutant overlords.
I wonder how much of this $7.8 billion figure is from bots gobbling up stocks of items to later resell at inflated prices. Apperently, according to another Slashdot article today, this is becoming a real problem and may well inflate the sales figure quite a bit.
I wonder what would happen if they made it a wiki instead of using github. Just let anyone edit the laws as they see fit, and get rid of congress.
I wonder what's best for the owner, though. Getting your car back in one piece without the thief, or getting it back with a thief inside who smashed the interior and windows trying to get out.
I don't know APK, and whether or not his complaints are valid, but his post might actuallty be unintentionally on topic. If reputation is a factor in this Chinese points program, and a few people with a grudge against you start "downmodding" you in this system, that can have pretty serious consequences. There was an article about a kid not being admitted at a school because his father's score was too low, for example.
Inventing is not just about having a concept like "maybe we can let a rocket land on its tail". It's also about actually making it work. Even the latest rocket being designed by NASA today, the SLS, is an expensive throwaway rocket. So what's the reason today?
Your advice to innovators seems to be "stop innovating, because you're just pretend engineers that are recycling old ideas". Fortunately, the actual innovators aren't listening to you and are actually pushing human society forward.
Have you ever invented anything?
Actually, he was also drinking quite a lot of whiskey during that interview. Few people seemed to mind, though.
SpaceX is not on the stock market. And Elon has contemplated taking Tesla off the stock market because apparently he felt it was holding them back by always forcing them to look at the short term.
As for recycling 100 year old technology: why haven't Boeing, Lockheed and the others done reusable rockets then, instead of wasting billions of dollars making expensive non-reusable rockets that are four times as expensive as even a non-reused Falcon 9? After all, it's not like this is brand new and exciting technology, they could have done so ages ago. With all the money NASA would have saved, they could have been on Mars by now. So why haven't they?
Or maybe the technology wasn't quite so simple and they all thought it was impossible.
Yes, my classical computer is perfectly capable of processing more than 2^17,000,000,000 different states just counting its RAM. I don't see what the problem is. Hell, I can easily draw one of many googols of possible states on a sheet of paper. Doesn't keep me from processing it.
I understand quantum computing is a lot harder than classcial, but if difficulty scaled with the number of states, the 72 qubit quantum computer made by Google, with almost 5 sextillion (10^21) states, ought to already be impossible. Seems to be working just fine regardless.
I, for one, welcome our new, Opportunity-imitating alien overlords.
This should be modded way up, even it it's an anonymous coward. The summary is indeed completely wrong, the radar actually works by comparing reflected signals to the signal that stayed home, thereby canceling background and jamming noise.
Doesn't matter how you do it. Entanglement cannot be used to send information. It can be used for encryption, though. Send entangled photons to two parties, they measure the photons so they each get the same, totally random key, and then they use that key to encrypt and decrypt a message sent via traditional means.
But doing something to the photons at one end and somehow directly affecting the other end is not possible. So you cannot tell whether or not a radar beam hit an aircraft by observing the entangled particles on the ground.
That's not how entanglement works. Maybe they are using some other kind of quantum effect. But entanglement actually does work instantaneously, faster than light. It just works in such a way that you cannot use it for communication. If you measure the particles at both ends, you will get the same result instantaneously (while you can prove that the decoherence occurred at that exact moment, not earlier, and therefore some kind of "information" must have traveled FTL) but you have no control over that result, and you cannot tell whether or not the other party made a measurement. So you cannot use it to send a message.
If this worked, it would allow faster than light communication. Send one half of an entangled pair of beams to someone (traveling at the speed of light, but that’s just the setup, not the actual communication), then the recipient disturbs the beam (or not), and the other party instantaneously notices. Nope, not possible.