Don't kid yourself. When it comes to AI universities are actually mostly just centres for incompetence and wasting big baskets of money..
My own project, begun in 1990 has been developing the theory for building a Strong AI since then - private research, no money no external backing.. With even a tiny bit of the kind of money the universities have wasted my project could have had a working machine by about 2005. The real problem with Strong AI is that it requires a lot of extrapolation and thinking well outside the box, plus application without many results over an extended number of years - something that most universities are terrible at by definition. (and most corporations are not much better at either)
One part of the equation I forgot to include. - The 'mean time between failure' for advanced type robots doing manual work is generally going to be less than a day, and for really tough jobs as little as a few minutes. That is one of the main reasons why the maintenance costs end up being so high. Its quite conceivable that each robot will require at least one full time technician just to keep it running. The reality though is that robots like Atlas are probably still at least ten years from entering the commercial marketplace, for all kinds of reasons. Again, the complexity should be compared to advanced aircraft not things like cars.
Simpler robots are a completely different story - their time has definitely come and they are already hitting jobs today. - But, at least for now they always have limits. The warehouse robot you mentioned still requires human hands on site when things go wrong... or if it drops a parcel. Autonomous vehicles will need people on board, if nothing else to deal with unforeseen circumstances and to stop them being stolen.
On the dual use / security side. We are already working on these things and solutions do exist. At least if my project ends up being the one that wins the Strong AI race - total security has been part of the design since about 1995. Other groups are or will be planning the same thing, probably in many different ways.. A Strong AI that is vulnerable to hacking just can not and will not be tolerated by people. Impregnable security is a requirement not an extra..:)
If you don't like Socialism - then go and try the lifestyle in somewhere with true unfettered capitalism - like Somalia. No pesky bureaucrats or police or justice system or teachers or medical help or anything else - unless you pay for it. Human life worth a dollar, human death slightly cheaper.
When you look at robots like Atlas don't compare it to car tech or PC tech, a better comparison is jet fighter or spacecraft tech. Such robots are astoundingly complex and pack a vast amount of wiring into a very cramped small complex space. The servos alone probably cost between about $2,000 to $15,000 each and a robot like Atlas would need about 20 minimum for main body and limb movement. Human like hands add another 20 to 30 smaller servos. The base cost for small production run models of Atlas (an older model) was $2 million per robot - and note that this $2million was probably a zero profit cost. In the future mass production could eventually reduce the costs to maybe $150,000 to 200,000 per robot. Outsourcing mass production to somewhere like China might reduce the costs to $50,000 per robot, but is likely to be almost impossible to do. - Like atomic bombs and ICBM's advanced robots and particularly Strong AI's are, or will be, very much dual use machines.. - Take a robot like Atlas, give it a gun, train it the right way, and its a soldier - and that's only the most basic obvious military use.. Ie its all very likely to be subject to military licencing. The big worst case situation for Strong AI is a widely spread machine that gets hacked and turned into a mass weapon against its users - security in these machines when or if they ever go public will be astoundingly high, like nothing you have ever seen before. 'Made in America' - or for the project I am working on 'Made in the UK'.
Oh Dear. The only problem with this research is that any black hole model that includes a central singularity already completely violates general relativity. For a black hole to have an external gravity field, energy has to escape beyond the edge of the black hole. At the outer event horizon the field has to cross an FTL barrier but this can be just about explained by gravitational red shifting. However inside the bulk of the black hole this barrier gets steeper and steeper until it reaches the centre. - To cross the FTL barrier from a black hole with a central singularity requires a speed that is almost FTL instantaneous, and there is no way that red shifting can explain that. The only explanation that works is an absolute FTL frame - which completely rules out general relativity as the primary theory of mechanics. The very existence of such a speed completely destroys the idea of relativity of simultaneity and instead requires an FTL simultaneity.. This also basically rules out dimensional time as a theory.. General relativity only applied to physics below the speed of light.
Munchausen by proxy is also fantastically easy to misdiagnose. With what I know about psychology and psychiatry I wouldn't even 100% trust it to be real. Neither field can claim to true understanding of how the brain works and both have a history of inventing 'quack' illnesses, and of massive misdiagnoses. I've also seen how good people like doctors or police can be at covering things up when they make a mistake - especially if it might cost them their careers or time in prison..
The irony of course is that Einstein did get pretty rich. He owned a few patents plus he was world famous and used his fame to tour the world giving lectures and talking in public. He also had a professorship and won various scientific awards.. He probably had a total wealth of a few million. ($10 to 20 million now)
There's a guy called Mark Zuckerberg, you would probably never guess he was rich if you just met him.. of course compared to Zuckerberg Einstein was poor..
Wind turbines are big oils solution to renewables. They make money three ways - subsidies, oil fired power stations as backup, and the actual electricity they sell.. And as a polluters bonus to power the world with wind turbines we would create an eco footprint so huge it almost makes even oil look clean..
1,000 x 200m high 5 to 10 MW wind turbines = 1 x large 1 GW nuclear power station. Except wind requires 50 x the territory + 500+ MW of gas or oil power backup capacity +
Ok you do have a point. You do have to remember that most PC laptops are made in relatively small numbers for each model, especially compared to companies like Apple.. So Apples service costs actually should be lower. Guess I was speaking from a position more of secondary sources than direct experience. I have read a lot of stories about the high costs of replacing iPhone batteries, and the large secondary market that has grown up does speak volumes..
tsotha I think you just made the 'list'. The list doesn't exist yet, but in ten or fifteen years a machine will parse this page and read your comment. You won't ever have to worry about whether abusing a sentient machine is real abuse - because you will probably never be allowed to own one..
One intelligent post followed by a mass of ignorance.. Jets burn fuel far hotter and gain far more power per unit of fuel than ordinary petrol engines. A car engine would melt at the temperatures a turbojet runs at. Modern bypass turbojets have something like 2.5 to 3 times the overall efficiency and 10 times the power to weight ratio of most petrol piston engines.. Electric today is nowhere near to competing with that.
You want one that looks, feels, and moves like a human? Yours for less than a million dollars. (maybe) We should be in production by 2028 at the latest.. (maybe)
My only point is that Apple makes far bigger mark ups than most other tech companies, and on top of that they already heavily overcharge for repairs, parts, and servicing. Creating a closed shop for servicing simply closes the loop further. In the future what Apple really want to do is to stop people from even doing stuff like changing out dud batteries, forcing them to use Apple servicing for that as well..
BTW : Sorry if previous post was a bit rambling - did write it at three in the morning..:)
I hate to tell you this but Apple are not a small bespoke company. If anything Apple phones are produced on even bigger scales than PCs - hence their per unit costs are about as low as they can be. This is especially true given the very limited number of different Apple models.
The common mass production rules don't really apply to silicon IC products anyway. The three biggest costs in chip production are the cost of the wafers, the purchase costs of the production systems, and the worker cost. Once you hit a base limit after setup and dev costs, further costs start to scale fairly linearly with production size.. So producing more doesn't reduce the cost per unit by that much. This is because IC plants work as many small production lines (or workstations) that run in parallel. In products like Apple the production costs are usually only a small part of the overall cost - about $100 to $200 out of say $500 to $600 for an iPhone 6.
Sadly not. Most of the callers are essentially just expendable 'robots' following their scripts, working for a few thin Rupees. The system decides who to call and when, and it follows its given list.. They have no power over it.. I'd rather be a sewer cleaner than a telemarketer..
I've had a few of these calls. I learned a way to wind them up till they almost explode. When they tell you you've got a virus ask for the software ID. They will just try to blaise by that but just keep asking, be calm and patient. Then explain you've got X number of machines and need to know which one is going wrong. One time I got a long silence then a noise like the guy had broken then he hung up.. As far as I'm concerned they are impersonating Microsoft and so are committing fraud.. they are criminals and deserve anything they get. (Need to send the Guantanamo Bay guys over to their call centres in India or wherever..)
Hate to tell you this but Apple products are also mass produced. They use the cheapest standard parts just like everyone else, they just sell them for more. They know that most people who buy Apple products are rich and stupid by definition - and inherently gullible..
Stage 1 - only use the cheapest parts so they are unreliable. Stage 2 - charge the earth. Stage 3 - rig the system so only authorised repairs will continue to work..
This lens is the kind of thing that might lead towards nano-scale optical computing. And is also a potential small step towards the holy grail of tech - molecular scale assemblers.. Its even made of roughly the right material - grapheme - not such a large step from diamond composites...
Slashdot Mirror
Don't kid yourself. When it comes to AI universities are actually mostly just centres for incompetence and wasting big baskets of money..
My own project, begun in 1990 has been developing the theory for building a Strong AI since then - private research, no money no external backing.. With even a tiny bit of the kind of money the universities have wasted my project could have had a working machine by about 2005. The real problem with Strong AI is that it requires a lot of extrapolation and thinking well outside the box, plus application without many results over an extended number of years - something that most universities are terrible at by definition. (and most corporations are not much better at either)
A law sending scientists to jail for 10 years for speaking without government permission? That's Kim Jong Un NK territory.
One part of the equation I forgot to include. - The 'mean time between failure' for advanced type robots doing manual work is generally going to be less than a day, and for really tough jobs as little as a few minutes. That is one of the main reasons why the maintenance costs end up being so high. Its quite conceivable that each robot will require at least one full time technician just to keep it running. The reality though is that robots like Atlas are probably still at least ten years from entering the commercial marketplace, for all kinds of reasons. Again, the complexity should be compared to advanced aircraft not things like cars.
Simpler robots are a completely different story - their time has definitely come and they are already hitting jobs today. - But, at least for now they always have limits. The warehouse robot you mentioned still requires human hands on site when things go wrong... or if it drops a parcel. Autonomous vehicles will need people on board, if nothing else to deal with unforeseen circumstances and to stop them being stolen.
On the dual use / security side. We are already working on these things and solutions do exist. At least if my project ends up being the one that wins the Strong AI race - total security has been part of the design since about 1995. Other groups are or will be planning the same thing, probably in many different ways.. A Strong AI that is vulnerable to hacking just can not and will not be tolerated by people. Impregnable security is a requirement not an extra.. :)
If you don't like Socialism - then go and try the lifestyle in somewhere with true unfettered capitalism - like Somalia. No pesky bureaucrats or police or justice system or teachers or medical help or anything else - unless you pay for it. Human life worth a dollar, human death slightly cheaper.
When you look at robots like Atlas don't compare it to car tech or PC tech, a better comparison is jet fighter or spacecraft tech. Such robots are astoundingly complex and pack a vast amount of wiring into a very cramped small complex space. The servos alone probably cost between about $2,000 to $15,000 each and a robot like Atlas would need about 20 minimum for main body and limb movement. Human like hands add another 20 to 30 smaller servos. The base cost for small production run models of Atlas (an older model) was $2 million per robot - and note that this $2million was probably a zero profit cost. In the future mass production could eventually reduce the costs to maybe $150,000 to 200,000 per robot. Outsourcing mass production to somewhere like China might reduce the costs to $50,000 per robot, but is likely to be almost impossible to do. -
Like atomic bombs and ICBM's advanced robots and particularly Strong AI's are, or will be, very much dual use machines.. - Take a robot like Atlas, give it a gun, train it the right way, and its a soldier - and that's only the most basic obvious military use.. Ie its all very likely to be subject to military licencing. The big worst case situation for Strong AI is a widely spread machine that gets hacked and turned into a mass weapon against its users - security in these machines when or if they ever go public will be astoundingly high, like nothing you have ever seen before. 'Made in America' - or for the project I am working on 'Made in the UK'.
If you call a ~ $500,000 to $1.5 million up front cost, plus $40,000 to $100,000 per year on maintenance costs 'minimum wage'.
And Bill turns back to the dark side yet again..
Typing Error : Last sentence, 'applied' should have been 'applies'.
Oh Dear. The only problem with this research is that any black hole model that includes a central singularity already completely violates general relativity. .. This also basically rules out dimensional time as a theory.. General relativity only applied to physics below the speed of light.
For a black hole to have an external gravity field, energy has to escape beyond the edge of the black hole. At the outer event horizon the field has to cross an FTL barrier but this can be just about explained by gravitational red shifting. However inside the bulk of the black hole this barrier gets steeper and steeper until it reaches the centre. - To cross the FTL barrier from a black hole with a central singularity requires a speed that is almost FTL instantaneous, and there is no way that red shifting can explain that.
The only explanation that works is an absolute FTL frame - which completely rules out general relativity as the primary theory of mechanics. The very existence of such a speed completely destroys the idea of relativity of simultaneity and instead requires an FTL simultaneity
Munchausen by proxy is also fantastically easy to misdiagnose. With what I know about psychology and psychiatry I wouldn't even 100% trust it to be real. Neither field can claim to true understanding of how the brain works and both have a history of inventing 'quack' illnesses, and of massive misdiagnoses. I've also seen how good people like doctors or police can be at covering things up when they make a mistake - especially if it might cost them their careers or time in prison..
The irony of course is that Einstein did get pretty rich. He owned a few patents plus he was world famous and used his fame to tour the world giving lectures and talking in public. He also had a professorship and won various scientific awards.. He probably had a total wealth of a few million. ($10 to 20 million now)
There's a guy called Mark Zuckerberg, you would probably never guess he was rich if you just met him.. of course compared to Zuckerberg Einstein was poor..
Figures. The DCMA people are Nazis and Nazism spreads like a disease..
Wind turbines are big oils solution to renewables. They make money three ways - subsidies, oil fired power stations as backup, and the actual electricity they sell.. And as a polluters bonus to power the world with wind turbines we would create an eco footprint so huge it almost makes even oil look clean..
1,000 x 200m high 5 to 10 MW wind turbines = 1 x large 1 GW nuclear power station.
Except wind requires 50 x the territory + 500+ MW of gas or oil power backup capacity +
Yep anonymous. Got to be a Neo Soviet infiltrator. Who else would sing the praises of the butcher Assad.
Ok you do have a point. You do have to remember that most PC laptops are made in relatively small numbers for each model, especially compared to companies like Apple.. So Apples service costs actually should be lower. Guess I was speaking from a position more of secondary sources than direct experience. I have read a lot of stories about the high costs of replacing iPhone batteries, and the large secondary market that has grown up does speak volumes..
tsotha I think you just made the 'list'. The list doesn't exist yet, but in ten or fifteen years a machine will parse this page and read your comment. You won't ever have to worry about whether abusing a sentient machine is real abuse - because you will probably never be allowed to own one..
One intelligent post followed by a mass of ignorance.. Jets burn fuel far hotter and gain far more power per unit of fuel than ordinary petrol engines. A car engine would melt at the temperatures a turbojet runs at. Modern bypass turbojets have something like 2.5 to 3 times the overall efficiency and 10 times the power to weight ratio of most petrol piston engines.. Electric today is nowhere near to competing with that.
You want one that looks, feels, and moves like a human? Yours for less than a million dollars. (maybe)
We should be in production by 2028 at the latest.. (maybe)
My only point is that Apple makes far bigger mark ups than most other tech companies, and on top of that they already heavily overcharge for repairs, parts, and servicing. Creating a closed shop for servicing simply closes the loop further.
In the future what Apple really want to do is to stop people from even doing stuff like changing out dud batteries, forcing them to use Apple servicing for that as well..
BTW : Sorry if previous post was a bit rambling - did write it at three in the morning.. :)
I hate to tell you this but Apple are not a small bespoke company. If anything Apple phones are produced on even bigger scales than PCs - hence their per unit costs are about as low as they can be. This is especially true given the very limited number of different Apple models.
The common mass production rules don't really apply to silicon IC products anyway. The three biggest costs in chip production are the cost of the wafers, the purchase costs of the production systems, and the worker cost. Once you hit a base limit after setup and dev costs, further costs start to scale fairly linearly with production size.. So producing more doesn't reduce the cost per unit by that much. This is because IC plants work as many small production lines (or workstations) that run in parallel.
In products like Apple the production costs are usually only a small part of the overall cost - about $100 to $200 out of say $500 to $600 for an iPhone 6.
Sadly not. Most of the callers are essentially just expendable 'robots' following their scripts, working for a few thin Rupees. The system decides who to call and when, and it follows its given list.. They have no power over it.. I'd rather be a sewer cleaner than a telemarketer..
I've had a few of these calls. I learned a way to wind them up till they almost explode. When they tell you you've got a virus ask for the software ID. They will just try to blaise by that but just keep asking, be calm and patient. Then explain you've got X number of machines and need to know which one is going wrong. One time I got a long silence then a noise like the guy had broken then he hung up..
As far as I'm concerned they are impersonating Microsoft and so are committing fraud.. they are criminals and deserve anything they get. (Need to send the Guantanamo Bay guys over to their call centres in India or wherever..)
Hate to tell you this but Apple products are also mass produced. They use the cheapest standard parts just like everyone else, they just sell them for more. They know that most people who buy Apple products are rich and stupid by definition - and inherently gullible..
And this is why only a fool buys Apple products.
Stage 1 - only use the cheapest parts so they are unreliable. Stage 2 - charge the earth.
Stage 3 - rig the system so only authorised repairs will continue to work..
Apple Fools !!!
Reposting This from inside the debate -
This lens is the kind of thing that might lead towards nano-scale optical computing. And is also a potential small step towards the holy grail of tech - molecular scale assemblers.. Its even made of roughly the right material - grapheme - not such a large step from diamond composites...