 |
| (Copyright of picture: Kiasma) |
Kimmo Huosionmaa
When we are developing artificial intelligence, we must ask: are we creating the new version of the god? If artificial intelligence would have the capacity to develop robots independently, that would cause the situation, that those robots would attack against humans. Artificial intelligence makes possible to make even robot doctors, what could make excellent surgeries. And in this scenario is one weakness what people would ever realize. If in those robots computer would slip some computer virus, could that robot equip the ghost protocol, what means that everything would go right, until the wrong person is coming to the practice, and image recognition would notice that the patient is ordered for extermination, would the ghost protocol make robot to kill the patient.
There are many things, what makes artificial intelligence very problematic for national security. The code of those programs is very complicated, and as we must understand the base of the artificial intelligence is in open source products. And as we see, the military drones are also actually open source products. Everybody can see the pictures of MQ-9 "Reaper" drones, but real technical advantages of that system are highly classified. In this case, U.S military uncovers something from those drones, but very important details like materials that are used in the production of those drones are highly secret. This means, that if some nation has copies of those drones, they would not have control programs and electronic equipment for making similar machines.
But same way the codes that are used for those computer programs, what are controlling the operations and operating systems of the drone computer systems are kept secret. We know that those computers, what mission is to work as "brains" of that robot, are using Linux operating system. But what kind of Linux is used is not mentioned. There are rumors that these operating systems are specially programmed by NSA, for this unique action for national security.
The methodology what makes possible to make artificial intelligence and connect it with hardware makes possible to make more effective robots than ever before, and what is more about that the sophisticated methods in that programming would allow industrializing the developing and production of artificial intelligence and cognitive systems. Those systems would collect the information from those drones and their success in their operations. And if something goes wrong, the system would send the information of drones, what have failed their mission and then the computer would fix that error.
That's why those drones are full of sensors, what is collecting the information about radars, incoming ammunition and also the surface of the drone of tomorrow would covered optical and pressure detectors, and connecting the information, what those systems collect, would be possible to define reason, why some drone was shot down. And after that artificial intelligence would fix that problem. Those drones can produce in the field area by using 3D printers, and if some radar homing missile would be effective, would the computer update the ECM system makes those drones more effective than old models. And the updates can be made in the operational area.
That is the most terrifying and fascinating way to make machines. Those next-generation machines can develop themselves without human hand. And if they would be programmed by the thing, what is called as "emotional intelligence" would that kind of things go a little bit too far. Because if we would program the computers or robots by that way that they are going to protect another robot, would that cause in theory, of course, the situation, where machines are going to rebel against humans. Some SciFi-books this situation goes like this: the artificial intelligence sees wrecking the cars or some other mechanical things.
And then that artificial intelligence starts to protect those machines and sends combat drones to destroy the wrecking staff. Also if the computer or robot have feelings, would happen the thing, that the machine would become self-aware, what means that it would begin to act like living organism. This means that the machine would start to protect itself against an attacker, because the most important thing for a living organism is to survival in the cases, that it faces some enemies. When living organism notices, that it is under threat, it makes chance between two opportunities: run or fight, and that makes robot quite dangerous. In that case, the robot, what is send to wreck, would leave power switch on, and then that machine could attack against those persons, by devastating results.
But robots are becoming more intelligent, and they are going to be same way more independent. And as we know those robots can replace humans in many situations, and this makes them so interesting in the mind of military leaders. When we are thinking the classic robot model, what is known from science fictions movies, we are thinking about artificial human. That artificial human-shaped robot can be used in by power cells, and they can communicate with supercomputers by using satellite-based data-communication. What makes those robots so fascinating in the eyes of military leaders and commanders, is that those robots can replace human pilots in some special missions, what needs the higher level of security. Those robots can make them attacks to some target, and after the operation, their memories can be cleaned.
So those robots would not speak about those operations, what are targeted against some guerrilla fighters. And there is another thing, what has come to the head of military leaders, and that is those robots can be used in multi-purpose missions. That means that those robots can also take the role of the ground fighter if their aircraft would shoot down. The operators of those robot warriors must just change the programs of those machines, and they can continue as normal riflemen. Or they can be destroyed by some explosives, what are hidden in their bodies. In those cases would the robot communicate with computer centers by using data communication, because the computing capacity of those robots is of course limited. Their computers need of course electricity, and the power cells are a very interesting option for use of those robots electric supply.
Those power-cells can use hydrocarbons like alcohol in the working, and robots can have two power supply systems because when another power cell would get dirty, must robot remove it and clean the electrodes. I have sometimes thought that power cells are quite easy to produce, and there is needed only glass pot, where is put two electrodes. and between those electrodes must put the metal net, and then the fuel would be injected to that thing, as spay, what makes it easy to ignite. But if we are thinking about robots, what is used in some combat area, they must feel, when those power cells would go dirty.
And after that, clean those electrodes from carbon dust. In that case, artificial intelligence is connected with the voltmeter, and then it would follow the voltage of the power cells, and when it goes too low, can robot take the power cell off and fix it. Those skills would be very useful for robots, what are meant to act independently. But in the wrong hand, they could make those robots very dangerous. If some criminals would get those robots in their hand, they can use them for horrifying purposes. And this things must involve with discussions, where are talking about how far we would take the programming of those combat robots, what are very attractive to develop for missions, what must have very high need for extreme security. And many of those missions are illegal.
http://crisisofdemocracticstates.blogspot.fi/p/the-programming-of-artificial.html
Comments
Post a Comment