Machine intelligence
After 70 or 80 years of rapid development, few people do not think that today’s computer is powerful. Machines can easily beat humans at their super-fast computing speeds, which allow complex numerical calculations to yield instantaneous results, and at their vast storage capacity, which can hold vast amounts of information.
On the other hand, if you ask yourself how intelligent is the machine? You might hesitate. While computers are powerful in some areas, they perform a lot of human intelligence worse than a three-year-old. For example, if you talk to one of the best robots in the world, it will sound childish after a few sentences. It is much more fun to talk to a three-year-old. So from this point of view, the current level of so-called machine intelligence is pretty low.
Weak artificial intelligence
The ultimate goal of artificial intelligence technology is to give machines the consciousness of thinking and make them work and think like human brains. In general, artificial intelligence can be divided into strong artificial intelligence and weak artificial intelligence according to whether they have self-awareness and independent thinking ability.
The type of ARTIFICIAL intelligence we often hear about today is weak artificial intelligence, which can only solve problems in a specific domain and is used more as a tool. Weak AI is built on big data and machine learning (including the current fad deep learning), which uses large amounts of calibrated data and algorithms to learn patterns of things. The training obtains a model parameter, then realizes the decision according to the model.
Strong artificial intelligence, on the other hand, refers to a variety of human abilities, such as independent thinking, self-awareness, seven emotions and six senses, reasoning and induction, and so on. It can be said that strong artificial intelligence has barely made progress and does not have the basis of theoretical engineering. It is more like a beautiful fantasy according to the actual situation at present.
Moore’s law
In 1965, Intel founder Gordon Moore discovered and explained that the number of transistors in integrated circuits (now known as chips) was doubling every year. He later revised his estimate, changing the cycle to two years, and eventually 18 months, by Intel CEO David House.
Thanks to Moore’s Law, the computing power of machines has increased exponentially over the past half century, and computers have become ever more powerful. Exponential change is such a frightening force that some people think that if computing power continues for just a few more decades, computers will become as powerful as humans.
But it is a fact that computers have long surpassed humans in arithmetic, but not necessarily in intelligence.
A mathematical expression
In many disciplines, such as physics, chemistry, geography and so on, mathematics can be used to describe how various systems work. If use the computer to realize the brain, then the operation of the computer is also a kind of system, if can use mathematics to express the operation can use the computer to realize the brain.
Since the brain is physically made of matter, there should be some kind of physical operation. And if computers have not yet achieved human behavior, then there should be some extension of existing mathematics and physics, and once those extensions are found, then we will have more powerful systems. If these principles could be described mathematically, it might be possible to achieve the functions of the human brain in a computer.
I think, therefore I am
So far there is no clear evidence that machines can think about this. We know next to nothing about the workings of the human brain. Nor do we know why philosophers say “I think, therefore I am.” Nor do we know how or why they came to this consciousness.
Everything is physics and mathematics
Some people are convinced that everything is physical and mathematical, and the brain is no exception, so long as it can figure out how things work it can be described mathematically and simulated in a computer. They were able to describe the brain physically and mathematically, and once they did, consciousness and thought naturally followed.
The brain is like an information processing device, and its behavior is data conversion and processing. It receives input from the senses, processes it through the brain, and then outputs it to the tissues.
conclusion
Can you make machines think? The question is whether you can simulate the human brain. At present, our understanding of the structure and mechanism of the brain is limited, so to speak, it is only theoretically possible. With the hundreds of millions of neurons in the human brain, and each neuron has thousands of connections to each other, the complexity is so high that we can’t simulate it right now. Even after the simulation, whether it will think on its own, whether it will generate consciousness on its own, these are not known.
————- Recommended reading ————
Summary of my open Source projects (Machine & Deep Learning, NLP, Network IO, AIML, mysql protocol, Chatbot)
Why to write “Analysis of Tomcat Kernel Design”
2018 summary data structure algorithms
2018 Summary machine learning
2018 Summary Java in Depth
2018 Summary of natural language processing
2018 Summary of deep learning
2018 summary JDK source code
2018 Summary Java concurrency Core
2018 summary reading passage
Talk to me, ask me questions:
Welcome to: Artificial intelligence, reading and feeling, talk about mathematics, distributed, machine learning, deep learning, natural language processing, algorithms and data structures, Java depth, Tomcat kernel and other related articles
