4 ms·

The Next Fear on A.I.: Hollywood’s Killer Robots Become the Military’s Tools

7 points by jbegley 3y ago

namaria 3y ago
Here is how things will play out: massive population movements will ensue when entire areas of the most densely populated countries on Earth become unlivable. Conflict zones with killbots will be established under one pretense or another to block such movements into rich countries. Exclusion zones will be reinforced by tactical nukes rendering transit zones deadly. You may not like it and downvote this. But we can already see the gears in motion. I have seen decades of warnings about imminent global warning. Yet the momentum has only increased. The Western Alliance has only a limited window of time where it can permanently block Chinese progress into military parity. They will act within it. The Taiwan conflict inevitability rhetoric is intensifying with each passing year.
rapjr9 3y ago
The military seems to have a lot of bogus arguments: “A core problem with A.I. in the military and in national security is how do you defend against attacks that are faster than human decision-making, and I think that issue is unresolved,” Mr. Schmidt said. “In other words, the missile is coming in so fast that there has to be an automatic response. What happens if it’s a false signal?” If it is coming in too fast and/or in too great a number to humanly counter then it is by definition not human. So what is the problem with destroying it? If the redundent sensors all malfunction that is a different problem that should be caught by long term testing (like the way we fix flaws in aircraft) especially before the system is armed. In general it seems like an automated defense system should be left off. It is only when current events make you think you'll need it should it be turned on, and that should be widely advertised so everyone knows it is on. If the argument is that this is a defense against a first strike nuclear attack, that's a laugh. Every major nuclear nation has had decades to smuggle in and hide nuclear weapons on the territory of their enemies. Stopping some missiles will not stop nuclear devastation. Even if you blow up all the nuclear missiles they will rain radioactive plutonium across the country and the world. Turning on an automated nuclear defense is a last ditch warning, nothing more. If you're defending against conventional missiles or hypersonic conventional missiles, some are likely to get through anyway (look at what Ukraine has been experiencing). In a war expect some losses, but also you know you are at war, hence you know when to turn on the automated defense system. As the article states, the Patriot system already can work this way. So is this really an unresolved issue or does the military not have a policy for when to turn on automated attack in a Patriot system? Jet fighter pilots already have to deal with attacks that are faster than human response times. They have automated systems for that. A bullet or shell can travel faster than human response time. A land mine detonates faster than human response time. This is not an unresolved issue, it is a common issue. The key question here is not response time, it is what does the AI actually do and how could it mess up? “I fought very hard to get a policy that if you have autonomous elements of weapons, you need a way of turning them off,” said Danny Hillis, a computer scientist who was a pioneer in parallel computers that were used for artificial intelligence. Mr. Hillis, who also served on the Defense Innovation Board, said that Pentagon officials pushed back, saying, “If we can turn them off, the enemy can turn them off, too.” Really? So the enemy can turn off all your jeeps that have ignition switches? Missiles have abort functionality, if the enemy figures it out they can abort all the missiles you fire. This is an old solved problem. Making sure an abort signal will actually work is an unsolved problem (witness the recent Starship launch where it almost did not). Misinformation and disinformation are already big problems even without AI. AI just makes it easier. It's not the AI that is the problem, it is knowing what is true, which has to do with people, not an AI. There are already many systems that verify truth, some good (digital signatures), some not so good (Social Security number). Work on the real problem, don't blame the AI because it makes it easier.
- rapjr9 3y ago
  Also, the argument "that if we don't do it our enemies will do it" is bogus. Biological weapons and chemical weapons have not proliferated and their use has been very rare. You might as well say "if we don't destroy them first they will destroy us, so we must immediately attack first" which gets you into Dr Strangelove territory. China did not expand their nuclear arsenal for decades, but now they have started building again. Why is that? It's because they saw the US and Russia doing it. So my counterargument is that if you build it, your enemies will also have to build it. Another counterargument is that if you build it, it will eventually find its way into your enemies hands. Russia's planes would not fly into Ukraine because Ukraine has old Russian air defense missiles. US Stinger missiles are in the hands of a lots of groups. In military strategy you have to plan to fight against your own weapons, because you have to assume the enemy has spies, or can just plain buy your weapons, or reverse engineer them. So if you build it, you have to assume your enemy will get it also.

7 points by jbegley 3y ago

Here is how things will play out: massive population movements will ensue when entire areas of the most densely populated countries on Earth become unlivable. Conflict zones with killbots will be established under one pretense or another to block such movements into rich countries. Exclusion zones will be reinforced by tactical nukes rendering transit zones deadly. You may not like it and downvote this. But we can already see the gears in motion. I have seen decades of warnings about imminent global warning. Yet the momentum has only increased. The Western Alliance has only a limited window of time where it can permanently block Chinese progress into military parity. They will act within it. The Taiwan conflict inevitability rhetoric is intensifying with each passing year.

rapjr9 3y ago

The military seems to have a lot of bogus arguments: “A core problem with A.I. in the military and in national security is how do you defend against attacks that are faster than human decision-making, and I think that issue is unresolved,” Mr. Schmidt said. “In other words, the missile is coming in so fast that there has to be an automatic response. What happens if it’s a false signal?” If it is coming in too fast and/or in too great a number to humanly counter then it is by definition not human. So what is the problem with destroying it? If the redundent sensors all malfunction that is a different problem that should be caught by long term testing (like the way we fix flaws in aircraft) especially before the system is armed. In general it seems like an automated defense system should be left off. It is only when current events make you think you'll need it should it be turned on, and that should be widely advertised so everyone knows it is on. If the argument is that this is a defense against a first strike nuclear attack, that's a laugh. Every major nuclear nation has had decades to smuggle in and hide nuclear weapons on the territory of their enemies. Stopping some missiles will not stop nuclear devastation. Even if you blow up all the nuclear missiles they will rain radioactive plutonium across the country and the world. Turning on an automated nuclear defense is a last ditch warning, nothing more. If you're defending against conventional missiles or hypersonic conventional missiles, some are likely to get through anyway (look at what Ukraine has been experiencing). In a war expect some losses, but also you know you are at war, hence you know when to turn on the automated defense system. As the article states, the Patriot system already can work this way. So is this really an unresolved issue or does the military not have a policy for when to turn on automated attack in a Patriot system? Jet fighter pilots already have to deal with attacks that are faster than human response times. They have automated systems for that. A bullet or shell can travel faster than human response time. A land mine detonates faster than human response time. This is not an unresolved issue, it is a common issue. The key question here is not response time, it is what does the AI actually do and how could it mess up? “I fought very hard to get a policy that if you have autonomous elements of weapons, you need a way of turning them off,” said Danny Hillis, a computer scientist who was a pioneer in parallel computers that were used for artificial intelligence. Mr. Hillis, who also served on the Defense Innovation Board, said that Pentagon officials pushed back, saying, “If we can turn them off, the enemy can turn them off, too.” Really? So the enemy can turn off all your jeeps that have ignition switches? Missiles have abort functionality, if the enemy figures it out they can abort all the missiles you fire. This is an old solved problem. Making sure an abort signal will actually work is an unsolved problem (witness the recent Starship launch where it almost did not). Misinformation and disinformation are already big problems even without AI. AI just makes it easier. It's not the AI that is the problem, it is knowing what is true, which has to do with people, not an AI. There are already many systems that verify truth, some good (digital signatures), some not so good (Social Security number). Work on the real problem, don't blame the AI because it makes it easier.

rapjr9 3y ago
Also, the argument "that if we don't do it our enemies will do it" is bogus. Biological weapons and chemical weapons have not proliferated and their use has been very rare. You might as well say "if we don't destroy them first they will destroy us, so we must immediately attack first" which gets you into Dr Strangelove territory. China did not expand their nuclear arsenal for decades, but now they have started building again. Why is that? It's because they saw the US and Russia doing it. So my counterargument is that if you build it, your enemies will also have to build it. Another counterargument is that if you build it, it will eventually find its way into your enemies hands. Russia's planes would not fly into Ukraine because Ukraine has old Russian air defense missiles. US Stinger missiles are in the hands of a lots of groups. In military strategy you have to plan to fight against your own weapons, because you have to assume the enemy has spies, or can just plain buy your weapons, or reverse engineer them. So if you build it, you have to assume your enemy will get it also.