he'll bilk lots of people out of tons of money and then retire as the company goes down in flames.
This, particularly since $35K doesn't seem like nearly enough money for what he's promising. That's enough to build the place, maybe, assuming he got the basic structure for next to nothing (which he probably did), but no way is that enough money to stock it with years of luxury supplies per person, and to keep it all maintained and updated for decades as we await the end of the world. Not to mention building and maintaining the airstrip that will be needed so the buyers have some prayer of getting to the place when the world is collapsing.
I suppose maybe it's like a timeshare, and there are hefty annual maintenance fees on top of the purchase price?
While that is all true, it ignores the real world...
The Samsung Galaxy line for example, comes with all kinds of stuff installed... like the Google Play store, GMail, Maps, etc...
None of which send anything to Google unless you use them, and none of which send anything privacy-impacting to Google unless you choose to give them permission to. You can set up a throwaway gmail account, install apps and navigate, etc., as much as you like and Google will never learn anything about you from your phone. Your phone will be less useful than if you provided it with all the information to make Google Now work, etc., but that's up to you. If you want, you can even disable all of the Google apps and install some other app store and some other set of apps.
Open source sounds nice, but what percentage of Android phone users run a truly clean copy of Android?
Are we talking about what Google does or what users do? Can you install Microsoft-free Windows at all?
You're conflating two issues: ability and interest. Android provides people the ability to modify their software in whatever way they like (at least on unlockable devices like the ones Google sells). The vast majority have no interest in doing that; they just want something that works, and Android plus the Google suite does an excellent job of that. Similarly, the vast majority of Windows users have no interest in modifying, or even understanding, their system... the difference is that Microsoft doesn't even provide the option.
Your claim was that Microsoft is starting to do what Google has been doing, but that's simply false. Windows has never been as open as Android, and that is not changing. Further, it appears that Microsoft has added user data-gathering which cannot be turned off (see previous articles about how there are controls but they appear not to work), which Google has never done.
Besides, Google has been doing it for years, and look how popular Android is.
Actually, Google works hard to make sure that users *do* own their own devices, or at least can. All Nexus devices are unlockable, and Google encourages OEMs to allow unlocking as well. Plus the whole open source thing.
Also, the common/. meme that Android reports on everything you do is simply false. Android the OS doesn't talk to Google at all. Google apps do, to the degree that you want to use them.
(Disclaimer: I'm a Google engineer, I work on Android, but I'm speaking only for myself, not for Google. Google has PR people for that; they pay me to write code.)
If a serial killer killed you during your Uber ride, how would you give them a bad review?
Uber would know you didn't arrive. I suppose the serial killer could deliver your phone to your destination...
Also, you could start calling 911 or taking other action with your phone as soon as you realized something was going wrong. That wasn't an option for taxis when the current regulations were set up.
Wrong, breeding for desired characteristic is an entirely different matter than what Monsanto is doing.
So, how do you feel about selective breeding processes that include drenching the organisms in radiation or mutagenic chemicals in order to dramatically increase the mutation rate? Nearly everything in your grocery store was bred via this method, which has been in use for at least a century, because it works really well. By massively increasing the mutation rate you can get your desired characteristics orders of magnitude faster than relying on natural mutations and cross-breeding.
If you're not okay with that method, then there's not much available for you to eat.
If you are okay with that method, can you explain how insertion of single gene to produce a desired effect is worse that thousands of random mutations, all of which are completely unknown outside of the immediately-observable phenotypic effects?
The fact is that humans have been doing various degrees of genetic engineering on our food crops for millenia, and massively increased it in the last couple of centuries (once Darwin explained how it worked). The methods of the last couple of decades are refinements which, if anything, should be dramatically safer than what came before, since the changes are smaller and better-controlled.
The taxi industry is regulated for very good reasons (one being safety)
I hear this all the time, but no one ever elaborates on what the reasons are. You said safety, but didn't say what the regulations are, how they are intended to affect safety and whether or not they really do.
One regulation that does make sense is the requirement that they carry commercial insurance policies. I think Uber has addressed that part (though I know some think Uber's solution inadequate).
As far as I can tell, the rest of the regulations are just an attempt to construct a functional reputation system in a context where little information is available to riders. By making it difficult and expensive for people to become cabbies, and relatively easy for them to lose that privilege, regulations ensure that only people who are serious about making taxi service a long-term business will do it. For exaple, this prevents J. Random Serial Killer from painting his car yellow and using it to pick up victims. Unless J. Random is also very wealthy, in which case he has lots of easier options. That's just one example, but the same line of reasoning applies to many other forms of abuse.
That all makes sense in a context where riders have no way to judge cabs other than by their appearance. But smartphones and the real-time, ubiquitous access to driver reputation databases they make available change the equation. Or so it seems to me.
Can anyone articulate precisely what other problems the regulations solve, and why the "rideshare" model (yes, I know it's not really ride sharing; let's discuss substantive issues, not quibble about naming) doesn't address them as well, or better? I'd like, for once, to have a conversation about this subject that goes beyond "Uber is exploitative and law-breaking!!!" and discusses the actual underlying issues. In what way, precisely, are cab regulations a better/safer/more efficient solution than ridesharing?
Obviously you can't invert md5, but if I hash my list, and you hash your list, and there is significant overlap, you can, to a reasonable but not 100% certainty, figure out which items on my list correspond to items on your list.
Depends on how it's done. For example, the advertiser could generate a bloom filter and provide that, rather than hashes of individual items on the list. Assuming the false positive rate was tuned correctly, you can use this method to arrange to provide very little information, while still generating the matches you want (plus some). Most advertisers wouldn't know how to tune the false positive rate appropriately, of course, but Google could tell them.
That's just off the top of my head, first glance at the problem. I suspect that there are even cleverer techniques that could be used, and while I don't know any details of how this system works, I do know Google engineers, and Google privacy design policies and procedures, and I'd be shocked if there were any obvious way to extract personally-identifiable information, in either direction.
(Disclaimer: I'm a Google engineer, but I'm speaking only for myself.)
So what you're saying is that all (or virtually all) campuses are gun free, so the fact this specific campus is gun free is pretty much meaningless.
No, actually. Several US states permit firearms on campuses. See the map at http://concealedcampus.org/sta... (hover over each state to see its rules).
...the way pedestrians act in Boston and New York: total chaos. People wander across the street randomly, and drivers are very aware that this is going to happen, so they slow down.
Interestingly, Boston and New York have very different pedestrian accident rates. New York has 1.52 pedestrian deaths per 100K, not much better than San Francisco's 1.70. Boston, though, has 0.79.
http://www-nrd.nhtsa.dot.gov/Pubs/811888.pdf
It's also worth pointing out that SF is actually safer for pedestrians than most big US cities. Boston appears to be the safest.
If they really want things to change, they should agree to work towards abolishing stupid patents---not to create semi-trusts that other companies have to fight.
Google has been spending tens of millions lobbying for patent reform, and only started to playing the patent game when it became clear that changing it wasn't going to work quickly enough -- though they haven't stopped trying to reform patents. The apparent contradiction has led some some pundits to question their motives, though I don't see that it's really a contradiction... the patent system is badly broken, but that doesn't mean Google can function in the industry as it is without playing the patent game. It's perfectly reasonable to play by rules you hate because that's what you have to do while simultaneously trying to change the rules.
Personally, I think software patents are a crock, but I'm listed as inventor on a few of them. I hate the game, but it is what it is so I play it while donating to organizations trying to change it. My rule is that I donate 50% of my patent bonuses to the EFF. I suppose if I were a better man I'd donate 100% (after taxes), but I do like to have some recompense for the effort I put into writing disclosures.
(Disclaimer: I'm a Google engineer, but I'm speaking for myself only, not for Google.)
Then, Google bought Softcard, and just killed the service. I guess they just have cash to burn.
Google didn't buy Softcard. The Softcard coalition shut down and handed their users over to Google to transition to Google Wallet rather than leave them with nothing.
It is also interesting to dive into those stats and you will notice a significant uptick of availability on weekends for north america. ISPs aren't the biggest offenders, nor is your home router, it is your company's routers and network that are the worst of the bunch here.
How did you find more detailed statistics?
I think a more likely explanation is the shift to mobile. Mobile networks tend to have much higher utilization of IPv6 and people are increasingly shifting to mobile-only for web browsing and especially for web searches. Weekends away from the office computers likely mean less time on keyboards and more time on phones.
Here is Google's official description of the feature: "If you don't want Google Chrome to save a record of what you visit and download, you can browse the web in incognito mode."
What if I don't want Google to save a record of what I visit and download?
Opt out. Google provides tools to enable it. https://support.google.com/ads.... Note that you can find that page by clicking "Privacy" at the bottom of http://www.google.com/ and then following the links embedded in the explanation of the issues.
And, yes, Google takes opt outs very seriously. A Google service found to be ignoring the opt outs would be considered to have a critical, don't-go-home-until-it's-fixed bug.
(I'm a Google engineer. I'm in no way an official spokesman, and speaking only for myself. But the comment on the priority of respecting opt outs reflects my personal experience of how such issues are handled.)
Yes they are. They're hiding from the cavemen with the nuclear weapons.
Any civilization with the technology required to travel interstellar distances has the ability to create weapons that make our biggest nukes look like fire crackers.
But by such stratagems, they could set up to that one person or a few people could flip a virtual switch, and the hack would be in place.
Very true. It's also not hard to construct an only slightly different sequence of events such that it was all a result of miscommunication, without anyone explicitly intending it. It will be interesting to see the root cause.
AI is not going to parse out every possible set of real world events which a driver may encounter. Google is finding that out now. The situations are too varied, too unpredictable (the technical term for this is "fucked up") and engage too many independent actors whose reactions are unknowable but critical.
Actually, I know some guys on the self-driving car team, and Google cars already handle just about everything safely. What they're focused on now isn't so much handling strange situations, but optimizing the car so it behaves like a human driver, to avoid confusing other human drivers -- or being taken advantage of by them.
OK that's the conceit of NN in a nutshell- just like a biological brain, so you said it. To me that's like saying a camera is like an eye. The brain is more than just neurons firing over synapses and reinforcing the ability to communicate across synpases. For example, nitrous oxide diffuses through the brain and is used in signalling. There are other things like that going on.
Meh. So there are some additional interconnections. Are those actually essential? It seems unlikely to me, but they could certainly be added if they are.
It may be the start of a good way to model the actual working of the brain.
Irrelevant. Oh, I suppose it may someday be relevant to neuroscientists whose goal is to understand the brain rather than to create useful systems. But for the people interested in being able to create automated systems that can be taught to make complex decisions effectively, what really matters is that it seems to work very well. Sure, the fact that we don't understand how they work means they may occasionally do insane things, but that's also true of complex decision-making algorithms we do understand (or think we do). And it's also true of living brains.
AI has a very very long and ignoble history of overhawking its wares, dating back to the 60s then the 80s then the 90s.. oh fuck it, every 10 or 15 years.
You say that as though everyone here isn't fully aware of it. But it's obvious, and it's common. It's in no way particular to AI. People are tremendously bad in general at predicting future technology for the -- rather obvious, actually -- simple reason that future technology will be built based on knowledge that we don't yet possess. You can't accurately predict the results of applying knowledge you don't yet have.
the idea that we're anywhere close to "the Singularity" which a lot of naive people believe, anywhere close at all is just not true.
You're wrong. And so are they.
The truth is that we have no idea how close we are to that point, and really won't have any idea until we're there, or until we prove that it's impossible.
However, none of that has anything to do with the topic at hand. Neural networks (biological or electronic) are almost certainly not the only way to construct the information flows underlying intelligence. It's also perfectly possible that our current NN models are inadequate for producing intelligence. So what? There are huge numbers of tasks for which we don't need general intelligence. All we need is a good automated decisionmaker which makes the right decisions and doesn't cost too much to build.
That is where NNs are awesome. How many engineer-weeks of effort would it take to produce an algorithm that, fed only the raw pixel data from the screen, can play a video game effectively? With Google's DeepMind NN, it takes one week of computer time, with little or no human involvement at all.
Neural networks may or may not be useful in reaching toward general AI. But they absolutely are useful tools, enabling us to build useful automated systems now.
I give you 100 to 1 that their self driving cars FAIL as a general mode of transportation by the year 2050.
I'll take that action. How much will you put on it? Let's define the terms and work out the logistics. Also, I'd be fine with pulling in that year by a couple of decades. I'll bet that self-driving cars on Google's model (full self-driving, highways and in town, with vanishingly few situations the car can't handle) will SUCCEED by 2030.
In a nutshell, I think it's a disguised way of doing statistics. An iterative, on-analytical way. With neural nets, after it's trained, no one can tell why the neural net functions as it does and no one can tell you when the neural net will do something completely insane.
Just like training a biological brain. And yet, those seem to be somewhat useful.
Or get the city to improve their trash collection process to use fewer people and cost less money. The service in my town costs $15 per month for weekly pickup. No flies and maggots, no trash accumulation, and less than one-third the price. $50 per month is outrageous.
Providing unconnected figures is meaningless in a vacuum. How much do you think a dozen eggs and a quart of milk should cost? Should it cost the same in Atlanta as it does in Alaska, Hawaii, or a big dairy state like NY?
Sure, costs vary. But 3X, for the same service in roughly the same context (single-family house), indicates a problem.
I'll bet the price of a house where you live is half the price of downtown Atlanta too
Overpopulation on earth will not be solved by colonizing Mars.
No, but reducing the asteroid metric[*] would be. Obviously not initially, since the Mars colony would be very dependent on support from Earth for quite some time, but you have to start with a dependent colony to create an independent colony.
[*] The number of very large asteroids needed to strike a planet to wipe out humanity. Substitute with any other sort of extinction-level event if you like.
Slashdot Mirror
he'll bilk lots of people out of tons of money and then retire as the company goes down in flames.
This, particularly since $35K doesn't seem like nearly enough money for what he's promising. That's enough to build the place, maybe, assuming he got the basic structure for next to nothing (which he probably did), but no way is that enough money to stock it with years of luxury supplies per person, and to keep it all maintained and updated for decades as we await the end of the world. Not to mention building and maintaining the airstrip that will be needed so the buyers have some prayer of getting to the place when the world is collapsing.
I suppose maybe it's like a timeshare, and there are hefty annual maintenance fees on top of the purchase price?
While that is all true, it ignores the real world...
The Samsung Galaxy line for example, comes with all kinds of stuff installed... like the Google Play store, GMail, Maps, etc...
None of which send anything to Google unless you use them, and none of which send anything privacy-impacting to Google unless you choose to give them permission to. You can set up a throwaway gmail account, install apps and navigate, etc., as much as you like and Google will never learn anything about you from your phone. Your phone will be less useful than if you provided it with all the information to make Google Now work, etc., but that's up to you. If you want, you can even disable all of the Google apps and install some other app store and some other set of apps.
Open source sounds nice, but what percentage of Android phone users run a truly clean copy of Android?
Are we talking about what Google does or what users do? Can you install Microsoft-free Windows at all?
You're conflating two issues: ability and interest. Android provides people the ability to modify their software in whatever way they like (at least on unlockable devices like the ones Google sells). The vast majority have no interest in doing that; they just want something that works, and Android plus the Google suite does an excellent job of that. Similarly, the vast majority of Windows users have no interest in modifying, or even understanding, their system... the difference is that Microsoft doesn't even provide the option.
Your claim was that Microsoft is starting to do what Google has been doing, but that's simply false. Windows has never been as open as Android, and that is not changing. Further, it appears that Microsoft has added user data-gathering which cannot be turned off (see previous articles about how there are controls but they appear not to work), which Google has never done.
Besides, Google has been doing it for years, and look how popular Android is.
Actually, Google works hard to make sure that users *do* own their own devices, or at least can. All Nexus devices are unlockable, and Google encourages OEMs to allow unlocking as well. Plus the whole open source thing.
Also, the common /. meme that Android reports on everything you do is simply false. Android the OS doesn't talk to Google at all. Google apps do, to the degree that you want to use them.
(Disclaimer: I'm a Google engineer, I work on Android, but I'm speaking only for myself, not for Google. Google has PR people for that; they pay me to write code.)
If a serial killer killed you during your Uber ride, how would you give them a bad review?
Uber would know you didn't arrive. I suppose the serial killer could deliver your phone to your destination...
Also, you could start calling 911 or taking other action with your phone as soon as you realized something was going wrong. That wasn't an option for taxis when the current regulations were set up.
Wrong, breeding for desired characteristic is an entirely different matter than what Monsanto is doing.
So, how do you feel about selective breeding processes that include drenching the organisms in radiation or mutagenic chemicals in order to dramatically increase the mutation rate? Nearly everything in your grocery store was bred via this method, which has been in use for at least a century, because it works really well. By massively increasing the mutation rate you can get your desired characteristics orders of magnitude faster than relying on natural mutations and cross-breeding.
If you're not okay with that method, then there's not much available for you to eat.
If you are okay with that method, can you explain how insertion of single gene to produce a desired effect is worse that thousands of random mutations, all of which are completely unknown outside of the immediately-observable phenotypic effects?
The fact is that humans have been doing various degrees of genetic engineering on our food crops for millenia, and massively increased it in the last couple of centuries (once Darwin explained how it worked). The methods of the last couple of decades are refinements which, if anything, should be dramatically safer than what came before, since the changes are smaller and better-controlled.
The taxi industry is regulated for very good reasons (one being safety)
I hear this all the time, but no one ever elaborates on what the reasons are. You said safety, but didn't say what the regulations are, how they are intended to affect safety and whether or not they really do.
One regulation that does make sense is the requirement that they carry commercial insurance policies. I think Uber has addressed that part (though I know some think Uber's solution inadequate).
As far as I can tell, the rest of the regulations are just an attempt to construct a functional reputation system in a context where little information is available to riders. By making it difficult and expensive for people to become cabbies, and relatively easy for them to lose that privilege, regulations ensure that only people who are serious about making taxi service a long-term business will do it. For exaple, this prevents J. Random Serial Killer from painting his car yellow and using it to pick up victims. Unless J. Random is also very wealthy, in which case he has lots of easier options. That's just one example, but the same line of reasoning applies to many other forms of abuse.
That all makes sense in a context where riders have no way to judge cabs other than by their appearance. But smartphones and the real-time, ubiquitous access to driver reputation databases they make available change the equation. Or so it seems to me.
Can anyone articulate precisely what other problems the regulations solve, and why the "rideshare" model (yes, I know it's not really ride sharing; let's discuss substantive issues, not quibble about naming) doesn't address them as well, or better? I'd like, for once, to have a conversation about this subject that goes beyond "Uber is exploitative and law-breaking!!!" and discusses the actual underlying issues. In what way, precisely, are cab regulations a better/safer/more efficient solution than ridesharing?
Obviously you can't invert md5, but if I hash my list, and you hash your list, and there is significant overlap, you can, to a reasonable but not 100% certainty, figure out which items on my list correspond to items on your list.
Depends on how it's done. For example, the advertiser could generate a bloom filter and provide that, rather than hashes of individual items on the list. Assuming the false positive rate was tuned correctly, you can use this method to arrange to provide very little information, while still generating the matches you want (plus some). Most advertisers wouldn't know how to tune the false positive rate appropriately, of course, but Google could tell them.
That's just off the top of my head, first glance at the problem. I suspect that there are even cleverer techniques that could be used, and while I don't know any details of how this system works, I do know Google engineers, and Google privacy design policies and procedures, and I'd be shocked if there were any obvious way to extract personally-identifiable information, in either direction.
(Disclaimer: I'm a Google engineer, but I'm speaking only for myself.)
Frankly, college students are adults, they should be able to have guns on campus.
Not all college students are mature enough to own a gun responsibly.
Apparently the college students in Utah are. Campus carry has been legal for more than a decade. Number of student shooting rampages: zero.
So what you're saying is that all (or virtually all) campuses are gun free, so the fact this specific campus is gun free is pretty much meaningless.
No, actually. Several US states permit firearms on campuses. See the map at http://concealedcampus.org/sta... (hover over each state to see its rules).
...the way pedestrians act in Boston and New York: total chaos. People wander across the street randomly, and drivers are very aware that this is going to happen, so they slow down.
Interestingly, Boston and New York have very different pedestrian accident rates. New York has 1.52 pedestrian deaths per 100K, not much better than San Francisco's 1.70. Boston, though, has 0.79.
http://www-nrd.nhtsa.dot.gov/Pubs/811888.pdf
It's also worth pointing out that SF is actually safer for pedestrians than most big US cities. Boston appears to be the safest.
If they really want things to change, they should agree to work towards abolishing stupid patents---not to create semi-trusts that other companies have to fight.
Google has been spending tens of millions lobbying for patent reform, and only started to playing the patent game when it became clear that changing it wasn't going to work quickly enough -- though they haven't stopped trying to reform patents. The apparent contradiction has led some some pundits to question their motives, though I don't see that it's really a contradiction... the patent system is badly broken, but that doesn't mean Google can function in the industry as it is without playing the patent game. It's perfectly reasonable to play by rules you hate because that's what you have to do while simultaneously trying to change the rules.
Personally, I think software patents are a crock, but I'm listed as inventor on a few of them. I hate the game, but it is what it is so I play it while donating to organizations trying to change it. My rule is that I donate 50% of my patent bonuses to the EFF. I suppose if I were a better man I'd donate 100% (after taxes), but I do like to have some recompense for the effort I put into writing disclosures.
(Disclaimer: I'm a Google engineer, but I'm speaking for myself only, not for Google.)
in business, things are seldom given. Especially with two very healthy mutinationals. You bet your ass Google paid; bought; paid a "fine".
Not in this case.
Then, Google bought Softcard, and just killed the service. I guess they just have cash to burn.
Google didn't buy Softcard. The Softcard coalition shut down and handed their users over to Google to transition to Google Wallet rather than leave them with nothing.
I understood that the opt out Google offered was for showing of targeted ads, not for the tracking. Am I wrong?
There are separate controls for both, ads and analytics.
It is also interesting to dive into those stats and you will notice a significant uptick of availability on weekends for north america. ISPs aren't the biggest offenders, nor is your home router, it is your company's routers and network that are the worst of the bunch here.
How did you find more detailed statistics?
I think a more likely explanation is the shift to mobile. Mobile networks tend to have much higher utilization of IPv6 and people are increasingly shifting to mobile-only for web browsing and especially for web searches. Weekends away from the office computers likely mean less time on keyboards and more time on phones.
Here is Google's official description of the feature: "If you don't want Google Chrome to save a record of what you visit and download, you can browse the web in incognito mode."
What if I don't want Google to save a record of what I visit and download?
Opt out. Google provides tools to enable it. https://support.google.com/ads.... Note that you can find that page by clicking "Privacy" at the bottom of http://www.google.com/ and then following the links embedded in the explanation of the issues.
And, yes, Google takes opt outs very seriously. A Google service found to be ignoring the opt outs would be considered to have a critical, don't-go-home-until-it's-fixed bug.
(I'm a Google engineer. I'm in no way an official spokesman, and speaking only for myself. But the comment on the priority of respecting opt outs reflects my personal experience of how such issues are handled.)
Yes they are. They're hiding from the cavemen with the nuclear weapons.
Any civilization with the technology required to travel interstellar distances has the ability to create weapons that make our biggest nukes look like fire crackers.
But by such stratagems, they could set up to that one person or a few people could flip a virtual switch, and the hack would be in place.
Very true. It's also not hard to construct an only slightly different sequence of events such that it was all a result of miscommunication, without anyone explicitly intending it. It will be interesting to see the root cause.
AI is not going to parse out every possible set of real world events which a driver may encounter. Google is finding that out now. The situations are too varied, too unpredictable (the technical term for this is "fucked up") and engage too many independent actors whose reactions are unknowable but critical.
Actually, I know some guys on the self-driving car team, and Google cars already handle just about everything safely. What they're focused on now isn't so much handling strange situations, but optimizing the car so it behaves like a human driver, to avoid confusing other human drivers -- or being taken advantage of by them.
OK that's the conceit of NN in a nutshell- just like a biological brain, so you said it. To me that's like saying a camera is like an eye. The brain is more than just neurons firing over synapses and reinforcing the ability to communicate across synpases. For example, nitrous oxide diffuses through the brain and is used in signalling. There are other things like that going on.
Meh. So there are some additional interconnections. Are those actually essential? It seems unlikely to me, but they could certainly be added if they are.
It may be the start of a good way to model the actual working of the brain.
Irrelevant. Oh, I suppose it may someday be relevant to neuroscientists whose goal is to understand the brain rather than to create useful systems. But for the people interested in being able to create automated systems that can be taught to make complex decisions effectively, what really matters is that it seems to work very well. Sure, the fact that we don't understand how they work means they may occasionally do insane things, but that's also true of complex decision-making algorithms we do understand (or think we do). And it's also true of living brains.
AI has a very very long and ignoble history of overhawking its wares, dating back to the 60s then the 80s then the 90s.. oh fuck it, every 10 or 15 years.
You say that as though everyone here isn't fully aware of it. But it's obvious, and it's common. It's in no way particular to AI. People are tremendously bad in general at predicting future technology for the -- rather obvious, actually -- simple reason that future technology will be built based on knowledge that we don't yet possess. You can't accurately predict the results of applying knowledge you don't yet have.
the idea that we're anywhere close to "the Singularity" which a lot of naive people believe, anywhere close at all is just not true.
You're wrong. And so are they.
The truth is that we have no idea how close we are to that point, and really won't have any idea until we're there, or until we prove that it's impossible.
However, none of that has anything to do with the topic at hand. Neural networks (biological or electronic) are almost certainly not the only way to construct the information flows underlying intelligence. It's also perfectly possible that our current NN models are inadequate for producing intelligence. So what? There are huge numbers of tasks for which we don't need general intelligence. All we need is a good automated decisionmaker which makes the right decisions and doesn't cost too much to build.
That is where NNs are awesome. How many engineer-weeks of effort would it take to produce an algorithm that, fed only the raw pixel data from the screen, can play a video game effectively? With Google's DeepMind NN, it takes one week of computer time, with little or no human involvement at all.
Neural networks may or may not be useful in reaching toward general AI. But they absolutely are useful tools, enabling us to build useful automated systems now.
I give you 100 to 1 that their self driving cars FAIL as a general mode of transportation by the year 2050.
I'll take that action. How much will you put on it? Let's define the terms and work out the logistics. Also, I'd be fine with pulling in that year by a couple of decades. I'll bet that self-driving cars on Google's model (full self-driving, highways and in town, with vanishingly few situations the car can't handle) will SUCCEED by 2030.
In a nutshell, I think it's a disguised way of doing statistics. An iterative, on-analytical way. With neural nets, after it's trained, no one can tell why the neural net functions as it does and no one can tell you when the neural net will do something completely insane.
Just like training a biological brain. And yet, those seem to be somewhat useful.
Or get the city to improve their trash collection process to use fewer people and cost less money. The service in my town costs $15 per month for weekly pickup. No flies and maggots, no trash accumulation, and less than one-third the price. $50 per month is outrageous.
Providing unconnected figures is meaningless in a vacuum. How much do you think a dozen eggs and a quart of milk should cost? Should it cost the same in Atlanta as it does in Alaska, Hawaii, or a big dairy state like NY?
Sure, costs vary. But 3X, for the same service in roughly the same context (single-family house), indicates a problem.
I'll bet the price of a house where you live is half the price of downtown Atlanta too
No, housing prices are pretty comparable.
Overpopulation on earth will not be solved by colonizing Mars.
No, but reducing the asteroid metric[*] would be. Obviously not initially, since the Mars colony would be very dependent on support from Earth for quite some time, but you have to start with a dependent colony to create an independent colony.
[*] The number of very large asteroids needed to strike a planet to wipe out humanity. Substitute with any other sort of extinction-level event if you like.
But they're selling you to 3rd parties.
More precisely, they're selling space on your screen to third parties.
I'm pretty sure Sean can verify that I *was* in fact a Google employee at one point in time, without violating *his* NDA. :)
Yep. Terry used to work for Google.