From this day on, Navalli officially entered Close AI's laboratory and began research work on "NLMR4".

The goal of NLMR4 is to create a strong artificial intelligence that can evolve itself with similar human perception. Fundamentally, its origins are completely different from all previous AI-like programs, including "NLMR3".

AI algorithms, including "NLMR3", are all artificially trained.

Although neural network algorithm models developed using deep learning can create AI programs that seem to be able to talk to people, they are still fundamentally simple structures that rely on human input data and then output it.

Even if the integration process has become more and more complex due to the increasing use of computing power, it has even become a kind of "black box", and the explainability is getting worse and worse.

However, it is still not a real strong AI, it is not spontaneous and inspiring, and it requires humans to broaden its application fields one by one and lay them on one engineering pipeline after another.

At the end of the day, it's still a "tool".

And the goal of "NLMR4" is by no means such a "tool".

- This is something that Navalli has known since day one.

But after actually participating in it, he was shocked by the level of Close AI.

……

"Hello, Navaly."

"Hello, Lucy."

"It's a nice day, Navaly, but you don't seem to be in a good mood? Why? ”

“…… I'm mistaken, my face has always been like this. ”

"Ah, I'm sorry, I thought that was an unhappy expression - it looks like I still have a lot to learn."

“……”

Navalli's fingers tapped on the keyboard, and after more than half an hour of testing, he loosened his neck and left the test table.

And just above the giant screen, the last words left by the AI named "Lucy" in this conversation seem to be burned in his retina - "Nice to talk to you!" Navalli, let's continue to talk next time. ”

And at this moment, another male voice came from his ear:

"How do you feel about our Miss Lucy, Navalry?"

The sharp-eyed brown-haired man smiled and handed him a bottle of mineral water, and said gently.

The man's name was Svador, and he was the specialized head of the "NLMR4" laboratory.

Being able to become the highest-level laboratory leader in the highest-level Close AI is naturally a good strength for Svaldor.

According to Navalli's information from Sanders, Svaldor already had more than a dozen top meetings as early as his college days. His doctoral dissertation "Self-Inspired Research Based on Neural Network Algorithms" has become a near-milestone research in the field of NLP (Natural Language Processing), with nearly 10,000 citations so far.

And after Svador graduated and entered Close AI, he never published a paper again. There are many people who suspect that he has encountered the difficulties of the "academia-industry" transformation that many bigwigs will encounter in Close AI, but now that it looks like the opposite, it is the opposite - Svalatorore does not publish papers not because of a lack of results, but because the results are too advanced, so he wants to reach the end in one breath until he creates a truly strong AI.

However, after Navalry's arrival, instead of showing any contempt for this man with no academic qualifications and academic qualifications, he was full of curiosity and attention to Navalry.

"I've studied the program you provided carefully - it's so well conceived and perfect, I never imagined that there would be such a clever segmentation algorithm."

Svaldo looked at Navalli with a falcon-like gaze and asked again:

"What do you think of 'Lucy'?"

Navalli took the mineral water, did not drink it, thought for a moment, and then replied:

"That's the strongest AI I've ever seen, no matter what angle you ask it any question, you can get an answer from an ordinary person's point of view - 'seems to work', and it can use other software and tools in a human-like way without a special interface - to some extent, it is not much different from the 'strong AI' that ordinary people know."

Svaldo stared at Navalli with no praise on his face, because it was not the answer he wanted.

"But?"

He prompted Navali to continue.

"But...... Frame problems and symbolic grounding problems remain. ”

Navalli looked at him and replied.

"It's still difficult to understand facts that are connected by abstract concepts. It can solve both one problem and the other very well – but it can't grasp precisely what is built on these two questions and that can be understood and responded to naturally with the human senses and intellect. And any human being, once he understands the first two questions, can answer the third question without any prior knowledge.

“…… So, although Lucy can answer most of the questions, her knowledge still does not go beyond the boundaries of the knowledge source based on human annotation. Although I haven't seen the full picture of its algorithms and code, I can be sure that its intelligent structure is still far from a real 'strong AI'. ”

It was a very unceremonious judgment, and Svalado stared into his eyes, narrowing gradually.

Tighter and tighter, more and more dangerous.

But Navali stared at him blankly as if unconscious, not even a little sweat on his body.

Long.

A smile suddenly bloomed on Svaldo's face, shining like a curtain suddenly opened, and he grabbed Navalry's hands and laughed at him:

"No bad! Even better than the best answer I could have imagined! Navaly, welcome to the 'NLMR4 Lab'! ”

Navali had a stiff smile on his face.

"It's my pleasure."

He knew that he had passed the introductory test for the person in charge...... Or a test.

However, there was no sorrow or joy in his heart.

At this moment, there was only one faint thought in his mind:

"My plan...... By this time, it had finally officially begun. ”

……

After entering the "NLMR4 Lab", Nava spent three days learning about and familiarizing himself with the lab's past algorithms.

According to the information obtained, the laboratory has been secretly conducting research work for more than three years, and at least hundreds of top engineers have come and gone here to leave the results of their efforts.

With the breadth of those codebases and materials, even Navali took three full days to roughly absorb them, and establish a framework understanding of the current algorithm results, AI "Lucy", and then officially entered the working state.

- Of course, this speed, even Svalatore is a little dumbfounded, not to mention the reaction of other researchers. Svaldo's expectation was that it would take him at least half a month to start the initial work, and this was the result of his overestimation.

After officially entering the working state, Navalli showed his maverick side compared to other engineers.

He doesn't like to collaborate, he likes to be alone, he buries his head in research, and he never talks about his ideas in exchange meetings.

There was a lot of discontent among the other engineers, but all of that was silenced by Svaldor, who, like Sanders, was a genius tolerant man.

Of course, the tolerance soon paid off.

Ten days after Navalli entered the laboratory, he developed a new path, which greatly optimized Lucy's adaptability, increasing Lucy's mechanical synesthesia level several times;

Eighteen days after entering the laboratory, he successfully optimized Lucy's algorithm, which raised Lucy's ability to judge complex situations in a simulated environment to a higher level.

After nearly a month in the lab, Navalli further increased Lucy's natural module logic ability, allowing her to improve the level of systematic logic within abstract concepts for the "natural world" that cannot be directly touched outside of the computer......

With so many achievements, the other engineers were finally convinced, and Schvador and Sanders, who came to the laboratory from time to time to check on the situation, had a higher and higher opinion of him.

If at the beginning, the two felt that he was a top genius who was "expected to get the Turing Award", now, this evaluation has changed.

- "He will be the Einstein of computers...... Even higher. ”

Sanders said this when he reported the situation to Salman.

And this evaluation also made Salman think for half an hour in his office chair.

Then, he smiled.

"Very well, it seems that we should pay more attention to 'Mr. Genius'."

He smiled, got up from his desk, and instructed the secretary something.

So, that evening, Navalli was again invited to dinner with Salman.

I was too tired last night to update, so I went straight to bed.

Today, I squeezed in time to update a chapter in the afternoon, and then add another chapter in the evening.

(End of chapter)