"Don't be too impatient. We haven't even understood the artificial intelligence technology of the Sibadi civilization." Lin Qiuyue replied lightly in response to Zhang Ran's questions.

"The network topology images sent by advanced civilizations are mainly based on the connectivist neural network system. This is actually a black box-level technology, and we actually don't quite understand why this thing works."

"This may be a big threshold..."

“Connectivism?”

After Lin Qiuyue's detailed explanation, Zhang Ran finally figured out the current development status of artificial intelligence.

There are three major schools of artificial intelligence today: the first is symbolism; the second is connectivity; and the third is behaviorism.

"The artificial intelligence developed by the three schools is completely different."

Lin Qiuyue turned on the laptop, and the smooth screen reflected her face with a hint of girlishness. She has always been a neat and neat style.

Zhang Ran sat on the other side of the dining table and listened quietly to her explanation.

"The concept of symbolism is to use the assumptions of physical symbol system and the principle of limited rationality, and advocate building an artificial intelligence system with axioms and logical systems."

"Simply put, it is to write artificial intelligence using programming languages ​​such as if...else, while...do."

"You will understand this way." Zhang Ran ordered two cups of coffee.

This is easy to understand. Most robotic technologies, such as intelligent control machine tools, computer operating systems, automotive control systems, etc., are the products of symbolism.

But it is obvious that intelligence developed based on symbolism has almost approached the human level to reach the upper realm, but it still has not reached the expected level.

"Simply put, it cannot handle content beyond the library and is quite far from the goal of self-learning and self-evolution."

"So there are connectionism and behaviorism."

The main principle of connectivity is the connection mechanism and learning algorithm between neural networks and neural networks, which is the current research direction of major laboratories.

This idea advocates imitating human neurons and using the connection mechanism of neural networks to realize artificial intelligence.

The most typical connectionist artificial intelligence is the Go software "AlphaGo" developed using deep learning algorithms!

This artificial intelligence has relatively strong learning ability.

"For example, deep learning algorithms... are neural networks composed of many layers of neurons to achieve the function of machine learning."

"Relying on the perceptrons of various layers provided by neural networks, deep learning machines that can be implemented in software in terms of functionality have achieved machine learning's mastery of induction."

"Since the whole process needs to be completed by the machine itself, there will be a large number of incomprehensible black boxes in the middle."

Lin Qiuyue explained: "The chess software 'AlphaGo' is trained by playing games with yourself. Due to the high complexity, humans can no longer figure out how it calculates it."

"I almost understand." Zhang Ran took a sip of coffee and nodded.

"No, you don't understand." A smile appeared in her eyes: "Have you ever heard of the artificial intelligence that was taught badly during the Earth era, its name is Tay, and the gender set to be a female, a chatbot."

Zhang Ran shook his head: "What do you say?"

"When she first appeared, Tay was polite. She was lively and chatty. She could observe the nickname, gender, hobbies, and emotional status of the chat subject, and would play games with you, tell jokes to you, comment on your selfies, etc. She could also learn popular languages ​​on the Internet. Connectionist artificial intelligence is good at learning."

"But soon, Tay began to speak rudely and swear constantly under the guidance of netizens, and even involved racism and pornography. After all, there is a lot of related content on the Internet."

"Her developers, who don't know how to modify her, can only force it to close it. This is how connectivity brings black boxes and uncertainty."

"That's it..." Zhang Ran took another sip of coffee.

Lin Qiuyue said: "However, this is just the first step to the black box, and there is a larger black box!"

"Have you ever thought about what would happen if two chatbots who can learn, chat with themselves and learn by themselves?"

"what happens?"

"They talk hundreds of millions of times in one second. As they learn from each other, there will be a lot of inexplicable garbled codes that we cannot understand."

"But it seems to be understandable to these artificial intelligences... They formed their own language system in just a few hours, and humans cannot understand these languages ​​and can only think that they are a series of garbled codes."

"We have no other way except to close them. This is the risk of connectivity, there are too many black boxes inside."

"We don't know whether these newly invented languages ​​really make sense, only machines can do it."

Zhang Ran was stunned for a moment, and she felt goosebumps all over her body as she said it. The robot actually learned from herself and developed her own language, which sounded a bit weird.

"You mean, will there be a smart rebellion?"

"There is such a possibility...but for the moment, caring for intelligence is equivalent to worrying about nothing, and the learning energy level can be roughly manifested as level nine." She broke out her slender fingers, "First, imitating words and deeds. Second, knowing how to use knowledge. Third, knowing how to associate. Fourth, knowing how to analyze and refine abstraction. Fifth, pursuing laws and verifying true knowledge. Sixth, accumulating true knowledge and forming a knowledge system. Seventh, guiding practical practice. Eighth, foreseeing trends. Ninth, disseminating theory."

"Only by completing these nine points is the real strong artificial intelligence."

"The wisdom of the orangutan has completed the first three points."

"Our current level is still pursuing the first level."

"Okay...what do you say behaviorism?"

The last genre is behaviorism, which believes that "intelligence does not require knowledge, representation and reasoning". As long as it interacts in the real environment and continues to provide feedback, intelligent behavior will gradually evolve.

This is also a bottom-up process, typical representatives include evolutionary algorithms, multi-agents, ant colony algorithms, etc.

The Sibadi people's artificial intelligence is also largely based on "behaviorism".

This type of artificial intelligence is an understandable category and will not have a large number of black boxes like neural networks.

Lin Qiuyue finally sighed: "Let me say this, the first genre, symbolism, has encountered a performance wall, at least our human intelligence is not good. Relying on the program library alone, we cannot even develop weak artificial intelligence..."

"The second school, based on the connectivityism of neural networks, will have a large number of black boxes to develop strong intelligence that we don't know, but its potential is the highest. Advanced civilizations also give extremely high evaluations, but there are actually quite a lot of difficulties at the theoretical level."

"The third school, behaviorism, is the most accessible path at present, and may be able to develop weak artificial intelligence that is no higher than humans. The Sibadi civilization also has similar robots, which will not lose control. The fully automatic universal factory was developed based on this concept."

"The renewal of universal factories requires humans to do it themselves."

"Or... the relationship between the three ideologies is not so conflicting and exclusion, but rather a mutually beneficial snake ring model. Through such symbiosis and comprehensive execution, can it be upgraded through behavioral training again and again?"

Zhang Ran also sighed and took his last sip of coffee: "How do you feel that the mission of the earth is to create human beings; and the mission of our human beings is to create stronger artificial intelligence to replace ourselves?"

But no matter what, artificial intelligence technology must continue to develop, because the core of the meta-universe era is a fully automatic production system based on the Internet of Things. Just pursuing the entertainment performance of the virtual world is just picking up sesame seeds and throwing watermelons.
Chapter completed!