I agree with you that current results in AI are very impressive. However they are all achieved using supervised learning. Even reinforcement learning is a form of supervision. Strong AI requires mixes of semi-supervised and unsupervised learning, i.e. the system does not get all the clues before being able to generalize, and is able to learn by itself. We are not there yet.
Amazing. As if all of the scaremongering and propaganda was on the "remain" camp. I urge you to read the main newspapers in England, like the so ever popular Sun.
If GB wants to remain in the EEA, they will have to accept that any EEA member can come to Britain and work without a visa. They will have to pay into the system too, like all the members.
Guess what, GB, the EU is exactly what members states do with it. There is no supra-national system that forces it to go one way or the others. One member state, one vote on the EU council. Representation is proportional to population in the EU parliament.
Not the idiots, idiots are easy to fool. You mean the dangerous, power grabbing, selfish sociopathic but unfortunately intelligent bastards you mention.
The proposed new French law about working conditions is way more encompassing than the summary suggests. However it does say something about email, it says that *employee* have the *right* not to consult email once they go home. In other words is will become illegal for an employer to expect employees to read and respond to email after hours.
All that is required of employers is drafting of a "code of good conduct" regarding the use of email. Compliance with the law will be voluntary and no checks will be made.
That does sound reasonable, if you ask me. Email and other forms of communication are encroaching on our life.
I don't completely agree with you, because potentially truly driverless cars can eliminate a lot of problems like drunk driving, which is responsible for a lot of fatalities. Drunk driving is already illegal yet happens. Possibly there could exist a technology other than driverless car that would eliminate it, like some sort of compulsory breathalizer test combined with other anti-tampering methods, but that would not be easy.
Nick Bostrom writes profusely and he writes about things that many people find fascinating. His papers are easy to read. That doesn't mean he is a lame thinker.
I'm surprised that no one has brought up the Feynman argument yet. TL;DR version: simulating physics is tough, it can't be done on classical computers.
If this universe is simulated, it is a hell of a simulation.
This is a huge shortcut. It is actually very difficult to simulate quantum phenomena on a computer. You can read this highly cited paper. Also interesting is this commentary.
[Citation needed] I don't find a source corroborating what you are saying, basically drinking water in the US seems to be ok, if not perfect. The problems don't seem to be associated to big cities.
Not exactly. The reward function was not learned (score function if you prefer), as far as I can understand from the paper here. This seems logical since otherwise there would be no way to tell the Deep-Learning CN whether a high score or low score was desirable.
That's not quite true. AlphaGo is a combination of classical tree-based search (actually stochastic search because the tree is too big to traverse deterministically) for selecting the best moves, and deep-learning for deriving the cost function.
It is well known that a human brain can operate a (finite) Turing machine, albeit very slowly. There are Turing machine in cardboard form for teaching the way computers work, here for instance.
So the brain can definitely emulate a Turing machine. The reverse is very likely to be true, because it is possible to simulate finite collections of atoms to a very high degree of accuracy with a Turing machine, including quantum effects. So in theory we *could*, with enough resources, simulate a whole brain down to the atomic level and run it.
Neither proposition is practical though. AI research is precisely about emulating intelligence practically.
What you claim as research is simply company communication. They have no duty of public disclosure unlike NASA. There are many other cost factors in engine reuse than fuel cost, like others have said on this very thread.
SpaceX will (and is) doing far better privately than NASA has publicly.
Right now SpaceX is riding off the back of NASA. We'll see how this pan out in the long run. Private enterprise is not able to rewrite the laws of physics. NASA may be rife with corruption and top-heavy with management, its engineers are still top-notch, and they are skeptical as well.
Mocking is not the right term. They are skeptical because both ESA and NASA have looked at the economics and practicalities of doing the same thing, and rejected the idea.
Slashdot Mirror
What makes you think post-brexit UK will opt out of the open border policy?
...Rita is a pimp, right? I thought it was illegal.
The promised year of Linux on the desktop ! or close enough.
Enough said.
I agree with you that current results in AI are very impressive. However they are all achieved using supervised learning. Even reinforcement learning is a form of supervision. Strong AI requires mixes of semi-supervised and unsupervised learning, i.e. the system does not get all the clues before being able to generalize, and is able to learn by itself. We are not there yet.
Amazing. As if all of the scaremongering and propaganda was on the "remain" camp. I urge you to read the main newspapers in England, like the so ever popular Sun.
If GB wants to remain in the EEA, they will have to accept that any EEA member can come to Britain and work without a visa. They will have to pay into the system too, like all the members.
Guess what, GB, the EU is exactly what members states do with it. There is no supra-national system that forces it to go one way or the others. One member state, one vote on the EU council. Representation is proportional to population in the EU parliament.
Yes, democracy sucks sometimes.
Not the idiots, idiots are easy to fool. You mean the dangerous, power grabbing, selfish sociopathic but unfortunately intelligent bastards you mention.
Why modded Troll, this is absolutely true.
This will come but it might take a few centuries.
If a referendum were held in all the Southern States, whether or not they would like to leave the Union, I would be curious to see the result.
The proposed new French law about working conditions is way more encompassing than the summary suggests. However it does say something about email, it says that *employee* have the *right* not to consult email once they go home. In other words is will become illegal for an employer to expect employees to read and respond to email after hours.
All that is required of employers is drafting of a "code of good conduct" regarding the use of email. Compliance with the law will be voluntary and no checks will be made.
That does sound reasonable, if you ask me. Email and other forms of communication are encroaching on our life.
I don't completely agree with you, because potentially truly driverless cars can eliminate a lot of problems like drunk driving, which is responsible for a lot of fatalities. Drunk driving is already illegal yet happens. Possibly there could exist a technology other than driverless car that would eliminate it, like some sort of compulsory breathalizer test combined with other anti-tampering methods, but that would not be easy.
True, but as a research project. Now we are talking about commercial product. Big difference.
Don't worry, Elon Musk will do that for us.
Nick Bostrom writes profusely and he writes about things that many people find fascinating. His papers are easy to read. That doesn't mean he is a lame thinker.
Actually Bostrom doesn't say and does not believe that we are living in a simulation.
Since when?
I'm surprised that no one has brought up the Feynman argument yet. TL;DR version: simulating physics is tough, it can't be done on classical computers.
If this universe is simulated, it is a hell of a simulation.
This is a huge shortcut. It is actually very difficult to simulate quantum phenomena on a computer. You can read this highly cited paper. Also interesting is this commentary.
[Citation needed] I don't find a source corroborating what you are saying, basically drinking water in the US seems to be ok, if not perfect. The problems don't seem to be associated to big cities.
Not exactly. The reward function was not learned (score function if you prefer), as far as I can understand from the paper here. This seems logical since otherwise there would be no way to tell the Deep-Learning CN whether a high score or low score was desirable.
That's not quite true. AlphaGo is a combination of classical tree-based search (actually stochastic search because the tree is too big to traverse deterministically) for selecting the best moves, and deep-learning for deriving the cost function.
It is well known that a human brain can operate a (finite) Turing machine, albeit very slowly. There are Turing machine in cardboard form for teaching the way computers work, here for instance.
So the brain can definitely emulate a Turing machine. The reverse is very likely to be true, because it is possible to simulate finite collections of atoms to a very high degree of accuracy with a Turing machine, including quantum effects. So in theory we *could*, with enough resources, simulate a whole brain down to the atomic level and run it.
Neither proposition is practical though. AI research is precisely about emulating intelligence practically.
What you claim as research is simply company communication. They have no duty of public disclosure unlike NASA. There are many other cost factors in engine reuse than fuel cost, like others have said on this very thread.
Right now SpaceX is riding off the back of NASA. We'll see how this pan out in the long run. Private enterprise is not able to rewrite the laws of physics. NASA may be rife with corruption and top-heavy with management, its engineers are still top-notch, and they are skeptical as well.
Mocking is not the right term. They are skeptical because both ESA and NASA have looked at the economics and practicalities of doing the same thing, and rejected the idea.
Now if SpaceX do it, all power to them.