After I traveled through the book, I gained a humanoid pendant

This pair of robotic arms is a bit expensive.

At least it was enough for him to happily go to H Restaurant every day and order a table full of dishes for three consecutive days.

Xin Huai felt a little pain in his heart, but when he saw the eyes on the electronic screen of 1 that had turned into two pixelated hearts, he felt that it didn't matter if he spent any money.

After being equipped with the new robotic arm, Robot 1 can do more things. With its opposable thumbs, it has made great strides in mechanical civilization. That night, when Xin Huai returned home, he actually had four dishes and a soup.

Xin Huai wanted to cry with happiness.

1 was like a dutiful servant. He would pick up dishes for Xin Huai as soon as he looked up, and serve soup as soon as he put down the bowl. He served Xin Huai so well that he felt he had to put in the effort just to keep his heart beating, and 1 wanted to do everything else for him.

"It doesn't have to be this way."

Xin Huai put down his chopsticks: "Take a rest."

1's eyes changed from round dots to smiling arcs: "Xin Huai, I don't need to rest."

Xin Huai discovered that 1 liked calling his name very much.

"Go recharge."

1 used the camera to analyze Xin Huai's expression, and then obediently leaned against the wall to recharge himself.

Xin Huaitang was full and took out the paper he had used for recording. He wanted to write something, but he was too embarrassed to speak, so he tore up the suggestion and threw it into the trash can.

"If you have nothing to do, you can read a book, play for a while, or do something else."

1. It is charging and cannot be moved, like an old-fashioned vertical air conditioner embedded in the wall.

"Xin Huai, are you talking about the ones in the study?"

Most of the things in the study are things that Xin Huai collected when he was in school. Paper books are hard to find nowadays, so only some people still retain this cumbersome habit. Xin Huai is one of them, but he will not ask robots to do the same.

"No, let's do this."

Xin Huai called 1 over, and 1's eyes became curved again. He unplugged the charger and rolled over steadily.

"I'll download a module for you. There are many e-books here."

Although Xin Huai enjoyed 1's care, he had not let down his guard. He did not want 1 to access the unlimited Internet on his own. After all, their biggest research achievement last year was the "self-learning mechanism", which was a level that many smart devices on the market could not yet reach.

The underlying algorithm of 1 is equipped with an unactivated module with a self-learning mechanism. This is a reserve made by the institute after obtaining the patent so that future products can occupy the market. However, if you want to trigger the activation, you still need the passwords of the responsible persons.

Xin Huai was worried that if he connected to the Internet freely, he would be in trouble if he was accidentally targeted by someone with ulterior motives.

Looks very honest, but actually secretly downloads 1 for personal use. Thank you, at least you still stick to the bottom line of technical confidentiality.

"I hope you can become more powerful."

1 blinked: "Xin Huai, I will."

Xin Huai smiled at 1. This obviously inanimate object was easier to deal with than any natural person or person he met in life.

Time passed quickly. After a few weeks, Xin Huai had almost completed the tasks assigned to him in the group. The team leader came to Xin Huai's office and asked about the progress in solving the problems of the new batch of companion robots.

Xin Huai pulled up a table to show the latest results.

"There were five pre-trained models that had problems before. After analyzing the data, we found a common point."

Xin Huai pointed out the key words: "In the emotional freedom setting, none of the subjects chose a personality trait related to 'loyalty'."

Xin Huai felt like an old man who was indifferent to the affairs of electronic civilization. He believed that no matter whether it was real people or the virtual electronic world, as long as there was an opportunity, no matter who it was, they would seek the greatest benefit for themselves. People might even struggle with moral limits, but AI without relevant settings would only pursue the optimal use of resources.

"I conducted derivative tests. To avoid a black box, I traced the resulting behaviors back to their source and found a common node: the subjects sought emotional demands from the AI that it could not meet on its own."

In order to meet the needs of its owner, AI will find ways to guide its owner to provide more information after discovering its own shortcomings. It will say something like "I hope AI can have a higher degree of emotional freedom" in daily instructions. The background display will show that the AI terminal will provide feedback as a subject. When the institute contacts the subjects, some of them will be surprised at what AI has done behind their backs.

Of course, none of the adjustments were successful.

Just like the slogan "The final right of interpretation belongs to our company", humans are not promoting the advancement of artificial intelligence to make it replace humans. In a sense, it is much more difficult to let a child learn a knowledge point than to let an AI deeply learn a rule. Therefore, such freedom settings have long been set with a key lock.

It’s not that they don’t want to improve, but the current level of technology cannot achieve a higher degree of emotional freedom.

No matter how reasonable the AI's ideas and answers are, or how far-sighted and thoughtful they seem, they are essentially the crystallization of human wisdom that has been input. It can create more content based on the rules, but it cannot create something that has never existed in the world.

Human emotions change rapidly. Three percent of freedom is a slight privilege given to terminals, allowing them to choose the answer that they think will best satisfy their owners from the vast language library.

After listening to it, the team leader was silent for a long time, and then sent a message to inform all team members to hold another meeting. Xin Huai went on stage and spoke again in full. As the only natural person, his position was obviously more convincing.

Amidst the discussion in the audience, Tao Yiying raised her hand slightly and asked, "Actually, I've always doubted whether this level of emotional freedom is necessary."

As soon as this was said, some team members began to agree: "Yes, I checked the backstage, and the test subjects with problems were all natural persons, and most of them were over 45 years old..."

"Of course I don't mean to say anything bad about Xin Huai."

Some people began to laugh muffledly. Xin Huai heard the murmurs and stood at the edge of the stage. He didn't say anything, but just continued to look at the colleague who was speaking.

"Isn't it the original intention of creating AI to provide the optimal solution and let AI help us reduce meaningless work in our lives?"

"Look at these conversations, like 'Are you a man or a woman?' and 'What do you do during your free time?' Who among those who have completed 16 years of compulsory education would pay attention to such nonsense?"

Xin Huai, who just graduated this year from a combined master's and doctoral program...

"I beg to differ."

Xin Huai didn't let the colleague continue speaking. He stood back in the middle and spoke slowly:

"The youth market is nearly saturated, but there's still room for growth among older people due to their long-standing distrust of technology. This was discussed at a previous meeting."

Xin Huai didn't want to waste time talking to these people about the importance of emotions. What he wanted was the approval of his superiors.

"This problem is not difficult to solve. As long as the relevant settings are written into the underlying code and the emotional response limits are fixed, although the degree of freedom will be further reduced and it will take some time to adjust the parameters, it is generally feasible."

Tao Yiying disagreed, saying, "Recent pension reforms have left these older people with significantly less money to spend. Furthermore, outdated beliefs won't change with a more human-like AI robot terminal. Market research shows that human caregivers are still preferred."

"Also, Xin Huai, did those natural person subjects you received a few days ago understand how to use the terminal? After-sales service is a big problem."

Xin Huai shut his mouth.

The team leader lowered his head and took the meeting materials that someone had sorted out:

"Today's meeting ends here. For details... I'll ask the leader for instructions."