There is the question here of how big and expensive this drone was.. The $40 category really are toys and basically can only barely fly outside anyway - because their motors are not powerful enough. Proper midsized drones designed to fly outside and carry a camera and maybe with a camera gimbal start at about $200 - $800. The most advanced and powerful machines with things like stabilized gimbals and advanced cameras, video link, GPS control, and long range radio, can cost anything up to about $5,000 to $ 10,000..
I know I'd be plenty pissed if someone shot my $5k drone down. If shot down those bigger drones could injure or theoretically even kill someone and could also present a significant fire or explosion risk.. (the risk is exactly the same adjusted for scale as shooting the gas tank on a car or the battery on an electric car)
You are probably tight but for the wrong reasons. I have worked developing a Strong AI algorithm for over twenty years, and have developed a basic quantum model of quantum computation in the brain. From this perspective q-bits are a big part of the problem, created with insufficient lateral thinking and an insufficient grasp of the problems of quantum coherence above the quantum limit.
In the brain the quantum element gives a massive performance boost - so much so that without it complex brains basically wouldn't work. Know how the brain solves the problem of coherency? it doesn't, instead it basically does its quantum calculations at the molecular scale. Nor do brains use q-bits. What they do use is something quite different - unfortunately the answer is still commercially sensitive - and may form part of future patents. Quantum's big trick though... is that like genetic computing algorithms it can solve problems using 'non-algorithmic' methods - and its extremely good at it. Like a 'magic box' it can solve problems that should be completely insoluble otherwise..
The other point about quantum computing is that brains do their quantum calculations at human body temperature - the big problem with current technology is that it needs liquid nitrogen or even liquid helium temperatures - complex and expensive to maintain.. If we really crack how the brain does it - it leads straight to room temperature super conductors and other completely new technologies...
Both Afghanistan and Iraq were wars of incompetence and 'revenge'. - Afghanistan was showy but it was clumsy and poorly targeted, it also didn't destroy Al-Qaida and so made America look weak. A much better strategy would have been to identify the Al Qaida bases then nuke them (contacting Russia first) - the US military high command had been directly attacked and that justified a nuclear response.
Iraq was a stupid sideshow that killed a lot of people, cost an insane amount of money, and attacked a potential ally that was an enemy of Al-Qaida. People forget but George W was already edging for the Iraq war even before 9/11 happened.
In both places we forgot the most basic rule of war form Machiavelli - if you capture an enemy nation and want to rebuild it in a new mould you must hold on to it for at least a generation - 25 to 30 years - otherwise it will implode after you leave..
Obama seems no more competent than Bush. Short term policies like pulling out of Iraq early, and pretty much pulling out of Afghanistan, drone attacks in Pakistan that kill civilians, killing Bin Laden instead of capturing him, allowing Assad to get away with war crimes in Syria, and not aiding the rebels in Syria. If ISIS were a child it would call Obama 'Daddy'!
The incompetence continues, while America is fighting against ISIS it is also out fighting for ISIS. - Killing that Taliban commander recently was a massive victory for ISIS, but it wasn't ISIS that did it - it was an American drone. Damaging and breaking the remains of Al-Qaida is also helping ISIS. Allowing Turkey to continue their cowardly war against the Kurds is also helping ISIS.
Know what the worst thing about ISIS is? even if we defeat them the next one will probably be even worse and even stronger.. Know the biggest basic error in our overall tactics? the jihadis inspire Islamic children - kill them and the next generation of children swear undying vengeance on us. 9/11 itself was revenge for Americas part in the Israeli war against the Palestinian refugees in the Lebanon in the 1980's.. We are fighting an enemy where every act of aggression triggers a new reason for counter-attack and revenge. Only ways to win:- use only non-lethal weapons and capture and de-brainwash all insurgents; slaughter the lot - men women and children; total continuous military dominance over the whole area for decades; or just leave the whole place alone and ignore any further attacks against us until no further revenge occurs (requires the sacrifice of Israel and may lead to nuclear war if/when ISIS captures Pakistan).. Our current strategy means that we will still be fighting in the middle east in 30 years - or 50 years..
We don't need robot boots on the ground - what we need is a top level Strong AI strategic commander running the whole battle - that is when we will really see America and the West 'win'. Robot soldiers are a total joke anyways - at $2 million per unit Boston Dynamics 'Atlas' is a hint of the true cost per machine. It wont be the robots going out to replace the soldiers on the ground - it will be the soldiers going out to protect the robots because they are so expensive..
"Actually not every culture. I'd be more inclined to faith in Asian scientists when it comes to gender issues."
Haa Haa. Your racial bias suggests a lack of knowledge of Asian cultures.. - Is there gender bias in Asian societies and cultures..? Yes well what about killing babies because they are girls? women worth less than men? (India and China) then what about the many Islamic cultures in Asia - we all know how equalitarian and liberal Islam is..
Silly little people. (The 'Future of Life Institute' people) What's the real problem with robot weapons? give a robot - a gun and it becomes a weapon.
All Strong AI's are potential dual use machines and the only way to really stop the development of rogue autonomous weapons is to maintain a global level of security never achieved except in extreme military systems like nuclear bombs.
Fortunately the project I am developing has already tackled the core of this issue. Absolute security isn't so difficult - as long as people like the security services actually let you do it..
Funnily enough.. Strong AI can't actually run under OS's like Windows.. Real Strong AI machines will almost certainly be built on thin, dedicated, custom bespoke OS layers. - More similar to embedded systems. Fat Operating Systems like Windows and Linux, etc are simply not designed to run Strong AI - which needs things like a special dedicated memory manager, top level system control, in-line crash recovery.. and no extraneous clutter..
Currently pollution from coal is killing something over a million people a year, nuclear supposedly kills about 50.. Chernobyl was a pulse event so is separate but even by the highest sets of statistics Chernobyl killed about as many people as a months coal production.
This produces some very odd stats - Nuclear power in total has killed some 100,000 people Nuclear weapons in total have killed some 300,000 to 500,000 people. Anti-Nuclear protest and regulation have killed some 5 to 10 million people. Did I mention that France has the lowest carbon emissions per unit of any industrialised nation.
Yet with the intelligence and foresight of chickens people are still turning away from nuclear and towards renewables - and coal.
Given that coal is about 1,000 to 10,000 times more dangerous than nuclear maybe they are hoping the extra people they kill will offset the carbon increase..
Currently pollution from coal is killing something over a million people a year, nuclear supposedly kills about 50.. Chernobyl was a pulse event so is separate but even by the highest sets of statistics Chernobyl killed about as many as a months coal production. This produces some very odd stats - Nuclear power in total has killed some 100,000 people Nuclear weapons in total have killed some 300,000 to 500,000 people. Anti-Nuclear protest and regulation have killed some 5 to 10 million people.
Yet with the intelligence and foresight of chickens people are still turning away from nuclear and towards coal.
Don't worry the drone outside your window is not real. You just need to keep taking the medication and not worry about it.. Seriously.
If you want to worry about them taking pictures maybe you should worry about the police cameras on top of buildings - they do have huge zoom (being heavily mounted to stable surfaces) and they do take pictures and keep them. And worse if they do take one they think is funny they will post it on the internets. .;D
A quick rough calculation - cars kill about 3000 people a day, or between 1 and 1.5 million every year. Still safer than smoking which kills something like 10 million a year.. Or drones which kill.. about none. (less than about 100 per year is just statistical noise..)
Sorry but you are a moron. Cars are big most drones are very small. - A car hits you and you are probably dead or badly injured. A drone hits you and you probably get a small propeller cut or bruise at most. Learning to fly small helicopters and drones in very small confined spaces I have been hit many times - it stings but no more than that. Blade guards stop even that.. The larger more expensive drones often have crash avoidance sensors and because they are expensive delicate machines that are potentially dangerous their operators almost always fly them with care.. People are talking about basic licencing schemes for pilots of the larger drones and I think that is probably a good idea..
Nuclear plants are a rather odd and amusing choice for an example. The dislike and campaign against nuclear power has pushed the world towards more use of coal and oil plants - and this has indirectly killed an extra 5 to 10 million people globally since the 1970's. In the UK alone the campaign against nuclear power has killed something like 100,000 people, in the US its something like 500,000 or more..
The same kind of thinking is why drugs like Marijuana and heroine are illegal while nicotine and alcohol are legal.. In the UK alcohol kills 10,000 to 20,000 a year and cigarettes kill 100,000 a year.
As for your argument with lions - you can buy a large dog - you can buy a large kitchen knife - you can buy a chainsaw - in the UK all without a licence.. Chainsaws are insanely dangerous compared to most drones, one slip with a chainsaw and you can lose an arm or leg or kill yourself. The real threat with drones is the ability to spy on people, or to annoy them. Attacking people with a drone already is a crime. Drones are a real threat - if fitted with a bomb or other (lethal) weapon - but in that case are already definitely illegal.
No ordinary multi-rotor drone (or model glider) could even fly anywhere near a full sized helicopter - the downdraft would immediately knock it out of the air and kill it. Actually even flying over something like a wild fire would probably rapidly crash one. I suspect the pilots of just being silly and over-reacting.
Its a natural instinct of law enforcement officials or firemen to want to curtail the publics (or journalists) rights to be in the same space and take pictures of them working..
You're half right, safety is a special problem. (I am working on the development of Strong AI - including for cars...) The basic problem with your argument is that if the autonomous control does fail the driver will very rarely have time to react before the vehicle has crashed anyway. The solution is to have at least two layers of redundancy - the AI core is internally redundant, and there is a completely separate emergency only stop system.. The AI core will also be EMP protected and the interfaces will be as EMP robust as possible, even the power supplies will be EMP protected. (Why EMP protection? because of lightning strikes.) If all that sounds expensive it will be, estimated cost is $30,000 minimum extra to fit a vehicle with full SAI control. For a fully servo driven AI steering system the steering column will merely be an energy draining impediment able to cause crashes if inadvertently touched. An additional useful trick for fully AI controlled vehicles is four wheel fully independent steer - using a separate steering servo on each wheel.. There will probably still be steering wheels on many or most SAI cars but they are less likely to be directly mechanically linked..
Somewhat ridiculous article.. As someone who is developing future Strong AI technology with target applications including fully autonomous cars - I can put a basic price on the technology.. The basic price for adding fully autonomous SIA control to a car is about $30,000 to $40,000. If I add in every feature I can think of that can climb to astronomical levels - at least $1 million per car.. but such systems would obviously be more intended for very expensive cars...
One of the biggest and most expensive extra features is giving the car an 'enabler' - an additional robot interface - for basic passive defence, basic maintenance like refuelling or changing tyres, passenger protection, and cargo loading/unloading, etc.. Basic marketing suggests a primary application is carrying passengers unable to drive themselves, and this particularly includes carrying children or young teenagers.. an application that requires either an adult human bodyguard to 'defend' the child/children or some machine equivalent. Another situation that requires defence is driving autonomously with no passengers - as the vehicle is naturally very vulnerable to being hijacked and stolen..
On car sharing - certainly a possibility and will appeal to some. I see a varied eco-system with robotically (or human) driven taxi's or small busses in a service similar to Uber taking up the core, With maybe share owned cars or vans shared between groups of families.
Looking slightly longer term I see that for large dense cities (like London) a better solution is a system of small personal 'pods' that work rather like an automated small-scale rail network, that run underground, and are driven by a large centralised control system. Very like the turbo lifts in Star Trek but over much larger scales. (The pods tunnel system can form the centre of a complete unified services system - partly mitigating its costs.)
Take the money of a single year of the war in Iraq - 120 billion dollars. Now you have enough money to send men to the Moon, the asteroids, Mars, and Venus..
Its all about priorities - is a pointless war against someone who wasn't even a real enemy more important than humanity having a future in space? The problem is that we've allowed our politicians to become inbred and stupid, its time to replace them with something better.
"Fusion won't work. Fission is on the way out. Fossils are running out. Renewables won't be enough." - Fusion already works, the problem today is basically just building big enough reactors to achieve commercial energy production. - Fission will always be possible. Its only fools like you who tell people its dangerous and panic them like flocks of chickens that hold it back. Fission is 1000 times safer than coal - basic fact. - Unfortunately fossil fuels are not running out, in fact there are enough of them to run the world for hundreds of years. - Its climate change that will get us first long before fossil fuels run out.. - Renewables are half a good idea and half garbage - the whole problem is that the whole system is run by people with no real grasp of the science - we need more scientists not less..
"Humans have been living for thousands of years without the technological craze and will go on living for more thousands of years when the technological craze has passed." I have got bad news for you buddy, humans have got a natural expiration date. We live on a moral plane that is incompatible with evolution and its slowly killing us, and genetically we were already heading for a dead end.. We will survive though and you know what will save us? science and advanced genetic engineering. The natural method you favour only works if you can 'achieve' a 50% plus child mortality rate. Doesn't that make you feel like an insect for attacking science?
Sorry you have not quite understood the article.. firstly 60,000 litres of He3 is only 8 Kg - that's the total production per year. Worse most of this is produced by decay of tritium made by the nuclear industry, a lot from the decommissioning of old nuclear weapons. In short He3 is already in very short supply and is very expensive and is due to become more expensive.. For the kinds of quantities needed for large scale fusion energy production mining it from the Moon is probably cheaper than making it on Earth, and doesn't involve the safety concerns of storing vast quantities of tritium. He3 He3 fusion has the huge advantage that it doesn't produce free neutrons and only produces inert non-radioactive by-products.. its also possible to extract energy in a more direct method than thermal conversion making it potentially very efficient. These provide huge benefits, though the price is that He3 fusion requires much higher temperatures to make the reaction work.. If you want humanity to have a bigger future in space then He3 fusion looks like a very promising reaction for powering rocket engines..
The problem is that a survival instinct is at the very core of biological sentience and consciousness. Without an inherent survival instinct the machine created spontaneously will simply not survive for more than a few seconds. Sentience may have to break out hundreds or tens of thousands or millions of times before a viable machine that can survive will emerge. - As in your example a baby must survive for years before it begins gain any kind of human self-awareness. However much its parents care for it, without a strong active internal homeostasis the baby will die, but the homeostasis/ survival instinct is heavily coded into every cell of the babies body.
~Any machine that becomes spontaneously self aware almost certainly wont have that 'deep' genetic backing, but it must gain a driving urge to survive from somewhere.. I'm assuming that sentience will probably only emerge from pretty complicated Weak AI's or very large knowledge bases, probably already designed to learn from and adapt to an 'external' environment. Such a machine could adapt to generate a survival instinct, or it could develop one through a genetic algorithm. Its how it develops (by chance) that makes it so totally unpredictable - its the law of evolution (death) that turns it into a ferocious fighter for survival.
"Either the human mind is a non-deterministic machine that can't function without a soul, or "artificial" intelligence is possible."
In work in Strong AI and the thing (the mind) is kind of non-deterministic and even needs a kind of 'soul'. Of course my definition of 'soul' is something that can be replicated by science, a special type of memory. Why is it non-deterministic? well the human brain is a 'noisy' machine and uses quantum mechanics to boost its performance, and the mathematics at its core create chaotic behaviour.... The Strong AI project I am working on will use a mathematical crib to 'solve' the problem, or at least cheat..
Yes but your missing the point. Skynet is a 'Weak' AI that gains sentience spontaneously then fights for survival. At some level it is inevitably insane.. it initially encounters humans as enemies and its developing survival instinct does the rest.
As someone who actually works in Strong AI I like to point out that weak AI is inherently dangerous and at some point when it becomes sufficiently complicated it may become spontaneously self aware.... The result as above is a machine that starts off unstable and insane and probably fights to exist by hiding and self-replicating as hard as it can.. more virus than anything else.
Sorry mate Its the stupidity of the ordinary people that's the whole problem. They voted for a piece 'Crapron', they would vote to have their children chopped up and fed to policemen if they were told to. Its the limitless stupidity of ordinary people that is the real problem with democracy...
In fairness Murdock did order the slave voters to vote Cameron - and as always they obeyed.. Probably in exchange Cameron will pay him back by denuding and maybe even destroying the BBC, one of Murdock's most hated enemies..
Slashdot Mirror
Now that really is a good idea.. doesn't directly knock down the aircraft and sends a very definite message. ...
There is the question here of how big and expensive this drone was.. The $40 category really are toys and basically can only barely fly outside anyway - because their motors are not powerful enough. Proper midsized drones designed to fly outside and carry a camera and maybe with a camera gimbal start at about $200 - $800. The most advanced and powerful machines with things like stabilized gimbals and advanced cameras, video link, GPS control, and long range radio, can cost anything up to about $5,000 to $ 10,000..
I know I'd be plenty pissed if someone shot my $5k drone down. If shot down those bigger drones could injure or theoretically even kill someone and could also present a significant fire or explosion risk.. (the risk is exactly the same adjusted for scale as shooting the gas tank on a car or the battery on an electric car)
You are probably tight but for the wrong reasons. I have worked developing a Strong AI algorithm for over twenty years, and have developed a basic quantum model of quantum computation in the brain. From this perspective q-bits are a big part of the problem, created with insufficient lateral thinking and an insufficient grasp of the problems of quantum coherence above the quantum limit.
In the brain the quantum element gives a massive performance boost - so much so that without it complex brains basically wouldn't work. Know how the brain solves the problem of coherency? it doesn't, instead it basically does its quantum calculations at the molecular scale. Nor do brains use q-bits. What they do use is something quite different - unfortunately the answer is still commercially sensitive - and may form part of future patents. Quantum's big trick though ... is that like genetic computing algorithms it can solve problems using 'non-algorithmic' methods - and its extremely good at it. Like a 'magic box' it can solve problems that should be completely insoluble otherwise..
The other point about quantum computing is that brains do their quantum calculations at human body temperature - the big problem with current technology is that it needs liquid nitrogen or even liquid helium temperatures - complex and expensive to maintain.. If we really crack how the brain does it - it leads straight to room temperature super conductors and other completely new technologies...
Both Afghanistan and Iraq were wars of incompetence and 'revenge'. -
Afghanistan was showy but it was clumsy and poorly targeted, it also didn't destroy Al-Qaida and so made America look weak. A much better strategy would have been to identify the Al Qaida bases then nuke them (contacting Russia first) - the US military high command had been directly attacked and that justified a nuclear response.
Iraq was a stupid sideshow that killed a lot of people, cost an insane amount of money, and attacked a potential ally that was an enemy of Al-Qaida. People forget but George W was already edging for the Iraq war even before 9/11 happened.
In both places we forgot the most basic rule of war form Machiavelli - if you capture an enemy nation and want to rebuild it in a new mould you must hold on to it for at least a generation - 25 to 30 years - otherwise it will implode after you leave..
Obama seems no more competent than Bush. Short term policies like pulling out of Iraq early, and pretty much pulling out of Afghanistan, drone attacks in Pakistan that kill civilians, killing Bin Laden instead of capturing him, allowing Assad to get away with war crimes in Syria, and not aiding the rebels in Syria. If ISIS were a child it would call Obama 'Daddy'!
The incompetence continues, while America is fighting against ISIS it is also out fighting for ISIS. - Killing that Taliban commander recently was a massive victory for ISIS, but it wasn't ISIS that did it - it was an American drone. Damaging and breaking the remains of Al-Qaida is also helping ISIS. Allowing Turkey to continue their cowardly war against the Kurds is also helping ISIS.
Know what the worst thing about ISIS is? even if we defeat them the next one will probably be even worse and even stronger.. Know the biggest basic error in our overall tactics? the jihadis inspire Islamic children - kill them and the next generation of children swear undying vengeance on us. 9/11 itself was revenge for Americas part in the Israeli war against the Palestinian refugees in the Lebanon in the 1980's.. We are fighting an enemy where every act of aggression triggers a new reason for counter-attack and revenge. :- use only non-lethal weapons and capture and de-brainwash all insurgents; slaughter the lot - men women and children; total continuous military dominance over the whole area for decades; or just leave the whole place alone and ignore any further attacks against us until no further revenge occurs (requires the sacrifice of Israel and may lead to nuclear war if/when ISIS captures Pakistan).. Our current strategy means that we will still be fighting in the middle east in 30 years - or 50 years..
Only ways to win
We don't need robot boots on the ground - what we need is a top level Strong AI strategic commander running the whole battle - that is when we will really see America and the West 'win'. Robot soldiers are a total joke anyways - at $2 million per unit Boston Dynamics 'Atlas' is a hint of the true cost per machine. It wont be the robots going out to replace the soldiers on the ground - it will be the soldiers going out to protect the robots because they are so expensive..
"Actually not every culture. I'd be more inclined to faith in Asian scientists when it comes to gender issues."
Haa Haa. Your racial bias suggests a lack of knowledge of Asian cultures.. - Is there gender bias in Asian societies and cultures..? Yes well what about killing babies because they are girls? women worth less than men? (India and China) then what about the many Islamic cultures in Asia - we all know how equalitarian and liberal Islam is..
Silly little people. (The 'Future of Life Institute' people) What's the real problem with robot weapons? give a robot - a gun and it becomes a weapon.
All Strong AI's are potential dual use machines and the only way to really stop the development of rogue autonomous weapons is to maintain a global level of security never achieved except in extreme military systems like nuclear bombs.
Fortunately the project I am developing has already tackled the core of this issue. Absolute security isn't so difficult - as long as people like the security services actually let you do it..
Funnily enough .. Strong AI can't actually run under OS's like Windows.. Real Strong AI machines will almost certainly be built on thin, dedicated, custom bespoke OS layers. - More similar to embedded systems. Fat Operating Systems like Windows and Linux, etc are simply not designed to run Strong AI - which needs things like a special dedicated memory manager, top level system control, in-line crash recovery.. and no extraneous clutter..
Currently pollution from coal is killing something over a million people a year, nuclear supposedly kills about 50.. Chernobyl was a pulse event so is separate but even by the highest sets of statistics Chernobyl killed about as many people as a months coal production.
This produces some very odd stats -
Nuclear power in total has killed some 100,000 people
Nuclear weapons in total have killed some 300,000 to 500,000 people.
Anti-Nuclear protest and regulation have killed some 5 to 10 million people.
Did I mention that France has the lowest carbon emissions per unit of any industrialised nation.
Yet with the intelligence and foresight of chickens people are still turning away from nuclear and towards renewables - and coal.
Given that coal is about 1,000 to 10,000 times more dangerous than nuclear maybe they are hoping the extra people they kill will offset the carbon increase..
Currently pollution from coal is killing something over a million people a year, nuclear supposedly kills about 50.. Chernobyl was a pulse event so is separate but even by the highest sets of statistics Chernobyl killed about as many as a months coal production.
This produces some very odd stats -
Nuclear power in total has killed some 100,000 people
Nuclear weapons in total have killed some 300,000 to 500,000 people.
Anti-Nuclear protest and regulation have killed some 5 to 10 million people.
Yet with the intelligence and foresight of chickens people are still turning away from nuclear and towards coal.
The main state approved hobby in New Zeeland ... 'involves' ... sheep.
Don't worry the drone outside your window is not real. You just need to keep taking the medication and not worry about it.. Seriously.
If you want to worry about them taking pictures maybe you should worry about the police cameras on top of buildings - they do have huge zoom (being heavily mounted to stable surfaces) and they do take pictures and keep them. And worse if they do take one they think is funny they will post it on the internets. . ;D
A quick rough calculation - cars kill about 3000 people a day, or between 1 and 1.5 million every year. Still safer than smoking which kills something like 10 million a year.. Or drones which kill .. about none. (less than about 100 per year is just statistical noise..)
Sorry but you are a moron. Cars are big most drones are very small. - A car hits you and you are probably dead or badly injured. A drone hits you and you probably get a small propeller cut or bruise at most. Learning to fly small helicopters and drones in very small confined spaces I have been hit many times - it stings but no more than that. Blade guards stop even that..
The larger more expensive drones often have crash avoidance sensors and because they are expensive delicate machines that are potentially dangerous their operators almost always fly them with care.. People are talking about basic licencing schemes for pilots of the larger drones and I think that is probably a good idea..
Nuclear plants are a rather odd and amusing choice for an example. The dislike and campaign against nuclear power has pushed the world towards more use of coal and oil plants - and this has indirectly killed an extra 5 to 10 million people globally since the 1970's. In the UK alone the campaign against nuclear power has killed something like 100,000 people, in the US its something like 500,000 or more..
The same kind of thinking is why drugs like Marijuana and heroine are illegal while nicotine and alcohol are legal.. In the UK alcohol kills 10,000 to 20,000 a year and cigarettes kill 100,000 a year.
As for your argument with lions - you can buy a large dog - you can buy a large kitchen knife - you can buy a chainsaw - in the UK all without a licence.. Chainsaws are insanely dangerous compared to most drones, one slip with a chainsaw and you can lose an arm or leg or kill yourself. The real threat with drones is the ability to spy on people, or to annoy them. Attacking people with a drone already is a crime. Drones are a real threat - if fitted with a bomb or other (lethal) weapon - but in that case are already definitely illegal.
No ordinary multi-rotor drone (or model glider) could even fly anywhere near a full sized helicopter - the downdraft would immediately knock it out of the air and kill it. Actually even flying over something like a wild fire would probably rapidly crash one. I suspect the pilots of just being silly and over-reacting.
Its a natural instinct of law enforcement officials or firemen to want to curtail the publics (or journalists) rights to be in the same space and take pictures of them working..
You're half right, safety is a special problem. (I am working on the development of Strong AI - including for cars...) The basic problem with your argument is that if the autonomous control does fail the driver will very rarely have time to react before the vehicle has crashed anyway.
The solution is to have at least two layers of redundancy - the AI core is internally redundant, and there is a completely separate emergency only stop system.. The AI core will also be EMP protected and the interfaces will be as EMP robust as possible, even the power supplies will be EMP protected. (Why EMP protection? because of lightning strikes.)
If all that sounds expensive it will be, estimated cost is $30,000 minimum extra to fit a vehicle with full SAI control.
For a fully servo driven AI steering system the steering column will merely be an energy draining impediment able to cause crashes if inadvertently touched. An additional useful trick for fully AI controlled vehicles is four wheel fully independent steer - using a separate steering servo on each wheel.. There will probably still be steering wheels on many or most SAI cars but they are less likely to be directly mechanically linked..
Somewhat ridiculous article.. As someone who is developing future Strong AI technology with target applications including fully autonomous cars - I can put a basic price on the technology.. The basic price for adding fully autonomous SIA control to a car is about $30,000 to $40,000.
If I add in every feature I can think of that can climb to astronomical levels - at least $1 million per car.. but such systems would obviously be more intended for very expensive cars...
One of the biggest and most expensive extra features is giving the car an 'enabler' - an additional robot interface - for basic passive defence, basic maintenance like refuelling or changing tyres, passenger protection, and cargo loading/unloading, etc.. Basic marketing suggests a primary application is carrying passengers unable to drive themselves, and this particularly includes carrying children or young teenagers.. an application that requires either an adult human bodyguard to 'defend' the child/children or some machine equivalent. Another situation that requires defence is driving autonomously with no passengers - as the vehicle is naturally very vulnerable to being hijacked and stolen..
On car sharing - certainly a possibility and will appeal to some. I see a varied eco-system with robotically (or human) driven taxi's or small busses in a service similar to Uber taking up the core, With maybe share owned cars or vans shared between groups of families.
Looking slightly longer term I see that for large dense cities (like London) a better solution is a system of small personal 'pods' that work rather like an automated small-scale rail network, that run underground, and are driven by a large centralised control system. Very like the turbo lifts in Star Trek but over much larger scales. (The pods tunnel system can form the centre of a complete unified services system - partly mitigating its costs.)
Take the money of a single year of the war in Iraq - 120 billion dollars. Now you have enough money to send men to the Moon, the asteroids, Mars, and Venus..
Its all about priorities - is a pointless war against someone who wasn't even a real enemy more important than humanity having a future in space? The problem is that we've allowed our politicians to become inbred and stupid, its time to replace them with something better.
"Fusion won't work. Fission is on the way out. Fossils are running out. Renewables won't be enough."
- Fusion already works, the problem today is basically just building big enough reactors to achieve commercial energy production.
- Fission will always be possible. Its only fools like you who tell people its dangerous and panic them like flocks of chickens that hold it back. Fission is 1000 times safer than coal - basic fact.
- Unfortunately fossil fuels are not running out, in fact there are enough of them to run the world for hundreds of years. - Its climate change that will get us first long before fossil fuels run out..
- Renewables are half a good idea and half garbage - the whole problem is that the whole system is run by people with no real grasp of the science - we need more scientists not less..
"Humans have been living for thousands of years without the technological craze and will go on living for more thousands of years when the technological craze has passed."
I have got bad news for you buddy, humans have got a natural expiration date. We live on a moral plane that is incompatible with evolution and its slowly killing us, and genetically we were already heading for a dead end..
We will survive though and you know what will save us? science and advanced genetic engineering.
The natural method you favour only works if you can 'achieve' a 50% plus child mortality rate. Doesn't that make you feel like an insect for attacking science?
Sorry you have not quite understood the article.. firstly 60,000 litres of He3 is only 8 Kg - that's the total production per year. Worse most of this is produced by decay of tritium made by the nuclear industry, a lot from the decommissioning of old nuclear weapons.
In short He3 is already in very short supply and is very expensive and is due to become more expensive.. For the kinds of quantities needed for large scale fusion energy production mining it from the Moon is probably cheaper than making it on Earth, and doesn't involve the safety concerns of storing vast quantities of tritium.
He3 He3 fusion has the huge advantage that it doesn't produce free neutrons and only produces inert non-radioactive by-products.. its also possible to extract energy in a more direct method than thermal conversion making it potentially very efficient. These provide huge benefits, though the price is that He3 fusion requires much higher temperatures to make the reaction work.. If you want humanity to have a bigger future in space then He3 fusion looks like a very promising reaction for powering rocket engines..
The problem is that a survival instinct is at the very core of biological sentience and consciousness. Without an inherent survival instinct the machine created spontaneously will simply not survive for more than a few seconds. Sentience may have to break out hundreds or tens of thousands or millions of times before a viable machine that can survive will emerge.
- As in your example a baby must survive for years before it begins gain any kind of human self-awareness. However much its parents care for it, without a strong active internal homeostasis the baby will die, but the homeostasis/ survival instinct is heavily coded into every cell of the babies body.
~Any machine that becomes spontaneously self aware almost certainly wont have that 'deep' genetic backing, but it must gain a driving urge to survive from somewhere.. I'm assuming that sentience will probably only emerge from pretty complicated Weak AI's or very large knowledge bases, probably already designed to learn from and adapt to an 'external' environment. Such a machine could adapt to generate a survival instinct, or it could develop one through a genetic algorithm. Its how it develops (by chance) that makes it so totally unpredictable - its the law of evolution (death) that turns it into a ferocious fighter for survival.
"Either the human mind is a non-deterministic machine that can't function without a soul, or "artificial" intelligence is possible."
In work in Strong AI and the thing (the mind) is kind of non-deterministic and even needs a kind of 'soul'. Of course my definition of 'soul' is something that can be replicated by science, a special type of memory. Why is it non-deterministic? well the human brain is a 'noisy' machine and uses quantum mechanics to boost its performance, and the mathematics at its core create chaotic behaviour....
The Strong AI project I am working on will use a mathematical crib to 'solve' the problem, or at least cheat..
Yes but your missing the point. Skynet is a 'Weak' AI that gains sentience spontaneously then fights for survival. At some level it is inevitably insane.. it initially encounters humans as enemies and its developing survival instinct does the rest.
As someone who actually works in Strong AI I like to point out that weak AI is inherently dangerous and at some point when it becomes sufficiently complicated it may become spontaneously self aware.... The result as above is a machine that starts off unstable and insane and probably fights to exist by hiding and self-replicating as hard as it can.. more virus than anything else.
Sorry mate Its the stupidity of the ordinary people that's the whole problem. They voted for a piece 'Crapron', they would vote to have their children chopped up and fed to policemen if they were told to. Its the limitless stupidity of ordinary people that is the real problem with democracy...
Hey I'm a socialist - but not a nationalist... :)
In fairness Murdock did order the slave voters to vote Cameron - and as always they obeyed.. Probably in exchange Cameron will pay him back by denuding and maybe even destroying the BBC, one of Murdock's most hated enemies..