Vol 6 Chapter 2757: Artificial intelligence cannot replace humans

Remember [New] in a second! In fact, in addition to these several control methods, we also have another control method, which is the intelligent voice assistance system. Different from the general intelligent voice assistance system, the intelligent voice assistance system we installed on this heavy-duty mecha is more intelligent, or it is an artificial intelligence program.

At this point in the introduction, Wu Hao said, "Start the intelligent voice assistant system."

"The intelligent voice assistance system has been activated, hello driver, Xuanwu-1 is at your service." Soon a virtual female character appeared on the screen, she said.

Wu Hao smiled at the two and introduced: "The tone says it is an intelligent voice system, but it is actually an artificial intelligence assistant, which can also be regarded as the main body of the mech, or the second pilot or the co-pilot. , its function is actually very simple, it is to provide drivers with various services, including some data broadcast reminders, and some related suggestions and so on.

In addition, it can also control and control the entire mech according to the instructions given by the driver.

It can even temporarily take over the control of the mecha after the driver loses consciousness, or after obtaining the driver's authorization, so as to control the mecha to perform some preset tasks or emergency tasks. "

"Since this kind of artificial intelligence can already control mechs, why do you need to set up a driver and let people control it. Isn't artificial intelligence control better, and didn't you say that unmanned intelligence is not the trend of the future?"

Wu Hao shook his head with a smile: "Artificial intelligence can indeed help you accomplish few things, and it is true that the mecha cannot be degraded and controlled by an artificial intelligence system.

But that poses a problem, and it's not a matter of choice. In fact, artificial intelligence wants to perform some tasks perfectly, but let it make decisions, that may cause some problems. "

what is the problem? Hearing what Wu Hao said, Luo Kai also became interested and asked.

At that time, you need to attack and destroy the bandits and rescue the hostages. And the artificial intelligence may choose to directly blow up the mission because of subsequent cases, and wipe out the gangsters and hostages together.

For you, for any weaponry, danger and reliability come first. "

That also means that the artificial intelligence system must violate those rule systems, and all actions must be within those rule systems.

The relevant decisions it executes are the least choices that are calculated by the algorithm, and that choice is based on the pros and cons of the algorithm and the data in the model library.

So the criteria for judging things that you now formulate for artificial intelligence are based on your values ​​and related social rules, legal systems, and so on.

For example, when you are performing some operational tasks, such as a fire in a military warehouse, the best choice is to evacuate the surrounding people immediately, in order to avoid personal injury.

But if the artificial intelligence is changed, it can do that. It will refuse to directly blow up the house to destroy the opponent because there are no civilians in the house, and will choose a weak attack to destroy the enemy and rescue the civilians.

Because although it is ten thousand times or hundreds of thousands of times stupider than your human beings, it still has the emotions of your human beings, and it knows what courage is, what responsibility is, and what belief is.

And once that code of conduct is released, the mecha controlled by artificial intelligence will become extremely safe, because it is no longer concerned about civilians.

Seeing the absent-minded expressions of the seven people, Wu Hao went on to say: "The most complicated and direct way is not to push them directly, or use heavy weapons to directly kill the opponent.

That is the most complicated way, and also the most costly way. It is definitely necessary to use a weak attack to rescue the hostages, but the gain is not worth the loss, and even that kind of house attack will cause heavy casualties on your side.

Luo Kai and Pang Zhengqing nodded, thinking that there was nothing wrong with what Wu Hao said.

But human creativity is inexhaustible, often at some critical moments, artificial intelligence will only choose the optimal solution step by step, but you people will choose to take risks to create positive.

Of course, that case may be too appropriate, but in terms of choices and trade-offs, artificial intelligence will always be inferior to humans.

You humans cannot be made weaker by us, but artificial intelligence can't. "That example is not controversial, but whether it is controversial in actual combat, anyone will do it if it is changed.

After talking about that, Wu Hao paused, and then he said to the seven people: "You can't give an example that you want to think about. During the war, you encountered a group of enemies. The group of enemies guarded a house Among them, and in the house, except for the group of stubbornly resisting enemies, there is no owner of the house with a family of eight.

For the robot, the best choice is not the best choice, but for you people, the best choice is often the best choice.

Seeing that the seven people agreed with what I said, Wu Hao continued to speak: "But you people are the simplest and most deviant group of animals. Is there anyone who will tolerantly violate those rules, systems, and even laws. Rules and laws It's one thing, but it's another thing to actually implement it. There are very few problems involved.

This is still the same example, you change it. During a large-scale battle, a group of gangsters fled into the house, took a family of eight as hostages, and then resisted tenaciously.

That sounds reasonable, right? "

Speaking of that, Wu Hao did not stop, UU reading www.uukanshu.com but went on to say: "Except for those, in terms of creativity, artificial intelligence will always surpass and replace human beings .

Wu Hao smiled and said: "Although you have been developing artificial intelligence technology, you are also very afraid that the development of artificial intelligence technology is beyond human control, so you will formulate some rules for it, some red lines, such as the famous The Eight Laws of Robotics was later changed to the Seven Laws of Robotics.

War is peacetime, and it does not have a set of rules for war. But that set of principles of war is so compliant, at most it can be put in the open.

And that house happens to be under the main route of your retreat, so if you go around it, what should you do at that time. "

Artificial intelligence may choose that way, but you humans will, and you will try every means to put out the fire, so as to minimize the loss.

Yes, you can't modify its code of conduct so that it can't become like a human, but the battlefield situation changes rapidly, and you can't guarantee that there will be no similar cases, but the artificial intelligence still makes the same decision.

It is true that artificial intelligence cannot perform some complex tasks, but when it encounters tasks that are difficult to imagine, it can do it. Tolerance to abide by those rules and laws is not always suitable for those tasks, especially military tasks in war. "