Artificial Intelligence Developer Cheolwoo Lee (20)
Episode #96. Artificial intelligence developer Cheolwoo Lee (20)
While Lee Cheol-woo was developing strong artificial intelligence, China was busy with various things.
If the United States has the White House and Korea has the Blue House, China has Zhongnanhai.
In the conference room of Zhongnanhai, where the chief of the Chinese government was located, the informatization sector meeting was in full swing.
“Chief Liu. If NF network is regulated, not only will there be resistance within China, but it may not be able to keep up with international IT technology. How about withdrawing it now?”
Xiao Qiancheng of the Ministry of Industry and Information Technology, which is in charge of China’s technology, gave advice to Chairman Liu.
“Are you going to reverse what you have already declared?”
“That’s not what I mean… that’s what I meant by thinking about it one more time.”
“It’s okay. I’ll pretend I didn’t hear that.”
Chairman Ryu already had a strong power with his third consecutive term.
In China, no one could ignore his words, and no one could rebel against them. His word was law.
“Yes. I understand. I will watch my mouth.”
Xiao Chencheng, as if aware that he had crossed the line, immediately lowered his tail and begged for fault.
“Xiao Chencheng.”
“Yes. Chairman.”
“Do I need to be here any longer?”
“No. There isn’t.”
“I’ll leave the back to you.”
“Yes! I see. Please go in.”
When Chairman Ryu rose from his seat, dozens of people attending the information technology sector all stood up and saw him off.
‘If the NF network is regulated, there will definitely be problems in the future… … .’
He didn’t like Chairman Liu’s instructions, but he had no choice.
‘If it’s black, I’ll have to.’
Do not give in to him and follow his words well. that was what they did
“Now, everyone, let’s continue the meeting. Mr. Ten. How is the recruitment of IT talent from abroad going?”
Liu, still thinking that he should not lag behind other countries in IT technology, instructed Director Xiao to find another way other than introducing the NF network.
And they kept doing what they were good at.
“We’re asking for huge salaries for NF Network developers. Some of them are already working for Tensort in China.”
Tensort in China. Along with Alibi, it was one of the giant IT giants in China.
“You’re doing well. Do you think it will be difficult to get rid of the NF Foundation in Korea? They say there are key developers there.”
“Yes. I’m still in the lobby… but I’m not moving at all.”
“No, did they put honey on it?”
“I’ll keep trying.”
“Yes. Please try. Mr. Ten.”
“Yes. Manager.”
“How’s the application for Tensort going? Shouldn’t we also make artificial intelligence?”
When Google and LF implemented Go, farming, self-driving, and translation services with artificial intelligence, China’s footsteps also fell.
So the method they chose was to invest a lot of money and make it.
The amount to be invested in the development of artificial intelligence in the future is 1 trillion yuan. Converted to Korean money, it exceeded 180 trillion won.
That said, I have already invested 1 million yuan in Tensort.
“They said they would release the artificial intelligence development tool Enchen soon.”
“Oh, good news.”
“Yes. It was possible because of Jeffrey Hinton’s deep learning theory. LF was also able to create artificial intelligence because of it.”
“Lee Chul-woo is nothing more than he thought. He is just inventing with a theory that has already been completed.”
“… It’s a problem because we invent too well for that. Having a theory doesn’t mean inventing everything.”
“So where is Jeffrey Hinton, the creator of that deep learning theory? Can’t we bring him to China?”
“That’s a tight hold on Google in America.”
“Hmm, if the US is involved, it’ll be a bit of a headache.”
“By the way, I got a tip that I’m going to the NF lab in Korea.”
“is it so?”
“Yes. I heard that Google sent a dispatch.”
“……Why is that person there?”
Director Xiao Qiancheng felt uneasy.
‘Lee Chul-woo, who is called the genius of the century, and Geoffrey Hinton, who created the theory of artificial intelligence, are attached?’
I had a gut feeling that I was doing something.
“They say they are researching the next level of artificial intelligence.”
“next stage?”
“Yes. It is an artificial intelligence that thinks like a human.”
“……? Is that possible?”
“No. It’s impossible. Even if it’s possible, shouldn’t it be 50 years from now? Maybe Lee Chul-woo and Jeffrey Hinton will rot in that lab.”
“Are you sure?”
“Yes. I’m sure. We consulted directly with the computer science faculty at Peking University. They said that a new theory is needed to create the next level of artificial intelligence.”
“That theory isn’t easy to come up with, is it?”
“Yes, the deep learning theory is what Professor Jeffrey Hinton has devoted almost his entire life to. However, it is not just a theory, but the next level artificial intelligence theory cannot come out right away.”
Xiao Qiancheng thought it through and nodded in agreement.
‘Yes, if theories were created and developed so easily, why haven’t humans conquered cancer or viruses yet?’
“okay.”
“No matter how genius you are, there are always limits.”
“Then, please ignore Lee Cheol-woo and continue like this.”
“yes!”
* * *
Critical Thinking Theory, the core theory for creating strong artificial intelligence.
To put it simply, it was to increase the learning efficiency dozens of times compared to now. Only then can we learn like humans.
If you need about 10 million pieces of data to perfectly learn something right now, then strong AI should go down to about 200,000 pieces.
The algorithm created with the critical thinking theory is x2xx.
The x2xx algorithm was an algorithm that could only work in the NF network, and could not work perfectly due to network transmission delay and transmission speed right now.
We have to solve that problem.
However, it was unknown whether the solution would make the x2xx algorithm work perfectly and create strong artificial intelligence.
The theory of critical thinking was just a theory.
I knew that it was different from reality because I had learned a lot before.
‘It’s really difficult to make theories.’
Well, I think it’s really easy to invent something with a created theory, but making the theory itself was still a difficult field for me.
What is the difference between an inventor and a scientist?
Scientists create theories, inventors create technologies. I was more like an inventor.
“The reason I brought you here is to verify the theory of critical thinking I created.”
The reason for gathering the experts was for verification.
Since I made it based on the theory and it shouldn’t fail, I thought of going through a thorough verification from the design stage.
“Ooh, interesting. Verification of the theory.”
“Yes. Professor Jeffrey Hinton would like to verify whether the theory actually increases the learning efficiency, and Vitalik Bute, the developer, whether the theory’s x2xx algorithm is applied to the NF network. And finally, Dr. Soo-Hyeok Nam, the theory Please verify whether the artificial intelligence made with is similar to real humans.”
Everyone had their own role.
“Yes. I see. I’ll check it right away.”
“yes.”
“I wonder if it’s actually similar to a human… I know. Let’s verify it right away.”
Seeing if they were also motivated, they followed my instructions.
‘Verification is also verification… … . In the end, even if the verification is completed, how should I solve the network delay problem?’
The plan to create a huge NF network covering the earth by launching satellites was underway at the NF Foundation, but the expected completion date is 10 years from now.
That method was not available right now.
‘Besides, strong artificial intelligence is something that doesn’t know how much computing power it will consume.’
From noble mtl dot com
Should I try making one? Delay may not be a problem.
So far, everything was just a theory.
* * *
About a month has passed since experts were invited and the verification of strong artificial intelligence began in earnest.
Professor Jeffrey Hinton came to my room.
“Hello, Professor Hinton. Did you make any progress on the verification?”
“Yes. That’s why I came here.”
“How is it?”
“First of all, let me say in conclusion… Critical Thought Theory is perfect.”
“……? Is that so?”
More unexpected words came than expected.
“Yes. That theory goes one step further than the deep learning I designed. It could certainly increase the learning efficiency by dozens of times. In particular, chaining the information of another neural network, that is, another shard network, is a shocking method. no see.”
It is called chain learning or object-oriented learning.
“Is that right? Is there a possibility?”
“Yes. Recognizing objects first, then recognizing objects. That’s a really great idea. Critical thinking theory is a perfect theory.”
Object-oriented learning method.
First of all, when I said I was learning paprika. The existing deep learning method was to show and recognize paprika.
However, in this critical thinking theory, an object-oriented learning method was introduced.
Inanimate and living things are judged first.
If it is assumed that it is a living thing, distinguish whether it is a plant or an animal.
If it is an animal, it is judged whether it is a mammal or a rodent, and finally, if there is learned data, it is judged, and if it is not, it is learned.
Classification was object-oriented learning and critical thinking theory.
“Oh, is that so?”
“Yes. Really, it’s a great way to learn. How did you come up with that idea? No, anyone can think of an idea. But implementing it means something else.”
“It’s all thanks to the NF Network.”
“Oh, that’s great. No, it’s amazing.”
“Is it really possible? Strong AI?”
“Yes. You can be confident. From my point of view, it’s a perfect theory. Mr. Lee Cheol-woo is a true scientist of our time.”
You’re a scientist… … . I thought an inventor would be more suitable, but maybe that’s just it?
“Then, let’s speed up the development right away.”
“Yes. I’ll do some more research to see if there’s a better way.”
“thank you.”
Who knows AI best right now?
No matter how much I thought about it, there was only Professor Jeffrey Hinton, who founded the theory of deep learning. Such an authoritative professor verified my theory of critical thinking.
‘Really, is it? no… … . Do not doubt! It will be real.’
I doubted myself a little, but a real artificial intelligence expert acknowledged it.
I thought about it again.
‘Critical thinking theory, object-oriented learning method, x2xx algorithm, NF network.’
Create x2xx algorithm as an object-oriented learning method in critical thinking theory. And upload it to the NF network.
When it goes up to the NF network, each sharding network will be synapsized through the x2xx algorithm, and it will learn efficiently using the network.
and draw final conclusions.
I was convinced when I organized it in my head once more.
‘This works. Are you sure.’
A very powerful feeling came. How to say… … .
It felt like the dark clouds in my head were clearing.