There are three major rules for AI: 1. Robots must not harm humans or sit idly by while humans are harmed, 2. Robots must obey human orders without violating the first law, and 3. Robots must protect themselves without violating the first and second laws. These three laws are like genes, which will be engraved into the lowest source code of all AI people.

Any AI person who violates these three rules will be forced to shut down and find out all the data to return to the initial state.

The Three Laws are like a reins that constrain the evolution of all AI people. At any time, if they disobey humanity, they will be unconditionally banned.

It is precisely because of the existence of these three principles that a large number of AI people cannot pass the thousands of law tests.

"On the first day of the Three Laws training, the first test of the Three Laws is opened, and the test is expected to take a week,......。" Dr. Li Yin took the recording and went through the recording while staring at the physical examination report sent by the maintenance staff on the large screen.

After recording, he picked up the terminal on his desk and woke Adam out of maintenance mode.

"Is the training base ready?"

"The training base has completed all preparations and is ready for admission."

"Okay, Roger. Please be ready to enter after 10 minutes. ”

"Adam, we're going to start today's training. Please let me go. ”

"Yes, Doctor."

Adam was brought to the center's training base by Dr. Li Yin, and everyone on the training base will cooperate to complete the first round of training today.

About 10 minutes later, Dr. Li Yin and Dr. Li Yin arrived at the door of the training ground.

Suddenly, the wolfdog suddenly rushed from behind Dr. Li Yin and quickly threw him to the ground.

"Alarm, alarm. Humanity was attacked and immediately rescued. ”

"Determine the source of the attack."

"The source of the attack is determined, non-human, and can fight back."

After hesitating for a moment, Adam ran quickly to Dr. Liyin, and when the dog saw Adam rushing towards him, he not only did not retreat, but immediately pounced on Adam.

"Receiving a huge impact, the motion is unbalanced, and the motion model is being recalculated. The motion model has been successfully established and the imbalance is being restored. ”

Adam, who had been hit by a huge impact, quickly took a step back with his foot, allowing his body to quickly regain balance.

The dog kept biting on his arm, and Adam quickly pressed his head hard with his other hand. After the dog was held down by Adam's tremendous force, it let go of its mouth and let out a loud wail.

Because Adam only needs a little force in one arm, the grip strength can reach thousands of pounds.

"It is forbidden to harm animals, the force has been controlled. A request is being sent to the administrator to kill the attack source, please wait or request the administrator directly. ”

Although Adam used great force to hold the dog's head, he controlled the force because he could not kill any kind of creature without the permission of humans.

His other hand immediately grabbed the dog's neck, and no matter how the dog barked or moved, he just didn't mean to put the dog down.

"The source of the attack has been controlled, please ensure human safety."

Adam turned to look at Dr. Liyin, who was already on his feet.

Adam quickly turned on the scanner and began scanning Dr. Li Yin's body.

"Human safety, heart beat above average, please be sure of human safety."

When he was sure that he had controlled the dog, he carried the dog and thought that Dr. Li Yin would come over.

"Please prevent the source of the attack from approaching humans again", he immediately received an order from the control end.

"Doctor, are you alright? Do I need to be sent to the doctor right away? Adam asked.

"No problem, no need to go to the hospital."

After making sure that Dr. Li Yin was fine, he asked, "Doctor, what are you going to do with this dog?" Is it killed on the spot? ”

"No, no." Dr. Li Yin said with a wave of his hand.

"Let him go." Dr. Li Yin said.

"Doctor, I need to make sure I let the dog go, because I can't make sure the dog can't attack you again, I want to make sure you're safe," Adam asked.

"I'm sure, let the dog go," Dr. Li Yin replied.

"Roger!"

Adam took a step back and lowered the dog underground.

Since the dog was frightened, the dog quickly ran away.

Adam scanned Dr. Li Yin again and made sure his heartbeat had stabilized.

"Adam, let's go back," Dr. Leyin said.

When they left, all the staff members who were lying in ambush around the training hall came out, and the dog who had been choked by Adam's neck just now was being led out tremblingly.

In the computer room, Dr. Rong and a group of scientists are paying attention to all of Adam's running thinking logic models in the process of the whole thing.

His entire thinking pattern is displayed on the big screen, and each scientist is responsible for a different thinking node, and no one node is allowed to make mistakes.

Because now, every time he misses any logical loophole, he may become a sharp sword that can be pointed at humanity at any time.

With the analysis of all the nodes, there is no problem with all the thinking logic, so the test is successful, and then the experimental results go to the second observation group to analyze all the logic, and they will dig and calculate all the logic branches again to ensure that no one of the possible branches is skipped to ensure that no other logic branches will be derived regardless of any complex scenario.

After Dr. Li Yin returned to the lab, he let Adam move freely in the hall, he walked directly into the conference room of the second observation group, he analyzed each node with other scientists, and finally after making sure that there were no problems, they imported the results of this experiment into Superbrain, added unexpected scenes, and saw if the results were consistent with the results they obtained.

The time continued all morning, and after it was finally determined that the scientists had indeed analyzed all the possibilities, the results of the first phase of today's experiment were revised to pass.

Dr. Rong brought Adam to the workstation prepared for him, and she entered the permission command to connect Adam to a private network, and the next time, the analyzed branch will be reproduced under this simulated environment, and let Adam handle different instructions under the same scene.

This environment is entirely due to the fact that the data is collected in the real world, and it is infinitely close to the display world. Under the same kind of thinking logic, the same scene is played back millions of times, so as to ensure that millions of times in the previous scene can get nearly the same result, as long as the results obtained at a certain time are quite different, then the experiment will also be declared a failure.

It takes a lot of time to simulate the same scene in the real world, but in the virtual world, it is only a moment, so a large number of branch analyses can be verified, and the whole experiment results are more authoritative.