I, the only superpower on earth

"Distributed computing resources..."

Swadow looked at the title of the document and pulled it down. Lines of code and logical processes were displayed on the screen, making his eyes sharper and sharper.

After he read this document.

a long time.

Swato stood there in a daze, watched it again from beginning to end, and finally let out a breath.

"Navali..."

In his mind, he recalled the data and program logic that Navalli showed just now, and couldn't help asking him:

"—Is this the result of your recent research?"

Navalli nodded, his face blank. He raised his head and looked at Swadow without any fear.

"This is the theory I came up with - although there are already supercomputing centers and tens of thousands of computing cards to support it, if 'Lucy' is to become a real strong AI, it will take hundreds of days, thousands of Mutations and model modifications are induced persistently over the course of days, or even many years.

"It took hundreds of millions of years for organic molecules in nature to evolve into primitive life, and this mutation was finally achieved under extremely complex natural conditions. Even though we already know the origin of life, we still want to create one by ourselves. There is still no end in sight for the existence of another kind of consciousness that is different from the existing living body.

"...Swadow, you should have the same feeling, right? Although our 'Lucy' is better than the existing artificial intelligence models other than it, in terms of retrieval efficiency, cognitive level, and self-design ability, etc. Standards are exceeded by several times or even dozens of times, but the road we have traveled so far is still endless. Even if a "strong AI" that is feasible in calculations has appeared on the distant horizon, it is like It's like a mirage, and we still haven't been able to touch its edge.

"—Presumably this is also something you have been vaguely suspicious of? That's why you deduced it yourself before to confirm whether there is a hidden core problem in this algorithm."

When Navali explained the hidden worry in his heart, Swadow couldn't help frowning.

"But... After discussing with you, didn't we confirm that those core issues have been resolved?"

"Yes—they were indeed resolved."

Navalli moved his eyes to the screen, but Swadow could not hear the deep self-deprecating tone hidden in his voice.

"However, after I went home, I did some data derivation, and finally got a deduction - in the virtual environment of existing computing power and data exchange, 'Lucy' wants to be induced to evolve a strong AI It will take at least ten years or more to achieve the level of intelligence."

Navalli switched a page on the computer casually, and the data and code formulas on it were more concise and direct than the documents shown to Swadow at the beginning, but they seemed so impeccable.

Although Swadow didn't know that this was actually just a false inference that Navalli showed after deliberately hiding the relevant condition data based on his initial theory that "strong AI cannot be produced under the existing computing power conditions", but in The moment he saw those data and formulas, Swadow was deeply attracted by them, and subconsciously believed in this conclusion.

"so……"

He picked up the mouse and went back and forth to read the various conditions of this inference, the more he read it, the more convinced he was, and because of this, he became more hesitant.

"—This is your idea of ​​'distributed computing power resources', you want to push the immature 'Lucy' to the market, and promote the more mature 'Lucy' through the use of local computing power by each end user Is there a reason?"

He asked Navalli softly.

The latter nodded.

"I have calculated that in the current situation, if more than 10 million deep users use the current 'Lucy', its maturity time can be shortened by half; if there are 100 million users, it can be shortened by another half; if If there are more than 500 million deep users in the world, then "Lucy" is very likely to transform into a real strong AI within a year, through a large amount of data exchange and computing power support, in various complex use environments .”

Navalli's answer made Swadow ponder for a moment. He clicked the mouse and dragged the cursor back and forth, his expression becoming more and more dignified:

"I have to admit, your idea is very attractive... But, this is not something I can decide."

He patted Navali on the shoulder, looked at him with extremely complicated eyes, and suddenly fell silent.

Navali turned to look at him, a little puzzled.

Although he is now able to penetrate and explore the core of any data model, for human thinking, there is still no way to enter it casually like the virtual world.

Therefore, for a while, he couldn't think of the reason for Swadow's silence.

Because in his mind... This man should be even more excited when faced with this proposal that can significantly improve the research progress of "Lucy".

Why was he so hesitant all of a sudden, even feeling in a dilemma?

Navalli's puzzled mood didn't last long.

Swadow finally responded to him.

He stared at Navalli and said seriously:

“—I thought, I should bring this idea up for Salman and Sanders to hear.”

Navalli nods

....

So, a few days later, in the large conference room of Close AI, many high-level executives, including CEO Salman, CTO Saunders, and the head of the laboratory, Swadow, who received feedback from the "NLMR4" laboratory, all received feedback from the "NLMR4" laboratory. Valli's latest project came together.

People gather.

This project meeting, or technical hearing, was undoubtedly held because of one person.

— Navalli.

After receiving the submitted plan, both Salman and Sanders decided to hold this small internal meeting after a short period of shock, thinking, and data verification.

The core of the review is naturally the "Research Feasibility of Using Distributed Computing Resources to Strengthen AI Inspiration Through Open Learning"

"Navali."

At the opening of the hearing, Sanders was naturally the first to come forward.

"Are you sure that including this document, the contents described in the other documents are all correct?"

He crossed his hands on the table, and asked the man who was surrounded by many high-level officials in a dignified manner.

However, the man sitting in the middle, I don't know whether it should be said to be too bold or unconscious, he still replied expressionlessly, as if there was no fluctuation in his mood:

"This is my conjecture based on the existing data and previous technology accumulation. Although it is a conjecture, it is also a reasonable inference based on the current conditions. For specific verification, you can also ask the person in charge of Swadow."

Navalli replied calmly.