Good afternoon, everyone was very pleased and honored to share some ideas with you at this grand event. In fact, at the time when Shen Hehe was summoned, when Cui Peng invited me, I wondered if it was necessary to discuss this thing. Afterwards, I really wanted to have some questions to discuss and discuss. I wanted to give an answer. Well, what's my answer? After a few days, I would like to sum up this sentence. It's a passion and calm. The best times.
In fact, at the time, Cui Peng told me that if you were to stand in opposition and talk about the bubble, I said no! I'm doing artificial intelligence, and then how do I think my direction has no future, so now I want to share with you why I think he's the best of times, but in these, but I think I might Talk more about the challenges and hope to share them with you.
First of all, let us take a look at what kind of artificial intelligence should be. In fact, on this two-day event, we have already said a lot. Many speakers have made very wonderful summaries, including our first two guests also gave very good. In fact, I would like to talk about this from a point. Thirty years ago, this man wrote a book 30 years ago. Of course I did not immediately see this book. This book was last year. There was a time when people especially liked to mention that robots would replace humans. I was amazed when this reappeared. When we were recommended by teacher Ma Shaoping of the task force at that time, he told me that there was such a book. Called the limit of artificial intelligence, what can computers do? 》
This book was written 30 years ago and was written in 1979. What I especially wanted to talk about in this book is actually a preface written by Professor Ma Xiwen. The preface was actually written in 1984. The book was published at the time of its final publication for 86 years. Actually, it has been a full 30 years. I read the preamble a lot, many times, for at least seven or eight times in almost two years. Every time I couldn't help but take it out and read it again, I personally thought it was a great benefit. I want to talk about it now. The three issues he mentioned in this preface.
The first question is whether the computer can complete a task that has hitherto been mainly accomplished by the human brain? The second question is whether the computer can do all this work? The third question is whether the computer can accomplish this kind of work like a brain? I think this is indeed a problem that we are concerned about. Let us first look at the first question. What I am very happy about is that the answer to this question 30 years ago is already certain. Look at that time when you read the book. Mentioned that machine translation, pattern recognition, problem solving, natural language processing, and so on, these fields are still areas where research and development has been in full swing and have made great progress. They do constitute one of the main contents of artificial intelligence.
But to see here we should not be too optimistic. The second question is whether the computer can do all this work. In fact, from the perspective of logical reasoning of science, the answer to this question is negative. The achievements made in the field of artificial intelligence may cause us to fall into some hallucinations. Why do we say so, in fact, Professor Lima did a very good job in this preface? The analysis, he said, if you want to let the computer solve a problem has three main prerequisites:
The first is that the problem that you solve must be formalized before it can be solved by the computer. However, how to formalize a problem, in particular, how to formally formalize the problem of the problem itself? It seems that we have fallen into a paradox here.
The second major problem is that this problem must be computable in order to establish an algorithm for it. This is an issue that we all have been very concerned with over the years. It is also one of the major waves of computer waves that have occurred over and over again over the years. The wave of artificial intelligence has formed one of the main driving forces for the new climax, but at the same time there is also a problem. How do we establish An algorithm to replace the human brain to find the algorithm? We can only say that the computer truly replaces human beings when we can achieve this realm in which the algorithm itself is not designed by people but designed by the computer. Unfortunately, at the moment we, at least for the moment, do not see this possibility.
The third problem is that you must have a reasonable degree of complexity in this issue. This is the reason why every climax in the past has fallen into a trough. In fact, this is often because of the initial success and the stagnation that followed. These stagnation are often limited by the computing power. Just now the guests also talked about this issue, so I think when we realize this problem, we will find that we are blindly optimistic.
So the third question mentioned earlier came. Can the computer accomplish these tasks like the brain? Everyone may think that the second problem has already been identified. The third problem is that we don’t think about it. In fact, this is not the case. Because you see, from then on to the present, from the earliest dark blue to the current alphago, etc. Computer programs have defeated human experts again and again, and we have solved many problems and even our search engine. They have exceeded the level that any of our experts can do. At this time we can see that On these tasks they can indeed work like brains and perform better than the brain. What we want to say in this case is actually, I hope our attitude towards computers and human brains is not what they should replace each other or how they should be. It is a system of mutual intergrowth and symbiosis. I think this symbiotic system is the real future of artificial intelligence and the opportunity for the most important development of artificial intelligence. It is also mentioned by so many experts in these two days. This is artificial. One of the future of smart development.
I’m telling everyone that, in fact, I want to tell everyone that we need to have a clear understanding of the current state of artificial intelligence . When you have a very clear idea of ​​what artificial intelligence can and cannot do, you won’t have What unrealistic illusions are not too disappointed and desperate about the difficulties they face, but can understand where the opportunities lie.
As the sole female guest of the forum, I am also a mother, so I think I can talk about my views on this issue from another perspective. That is the opportunity. I am doing artificial intelligence. When I have children, I tend to involuntarily carry my children to a certain level of analysis. I will think about what we should learn from people to help us do manual work. intelligent? What interesting things can we learn? So I want to start with the cognition of my three-year-old child and tell us a few stories to draw some of my opinions.
When my child was one year old, he wouldn't speak too many words at the time. I showed them picture books. We showed her many pictures, such as the moon and the bunny playing on the moon. Of course, we have this picture. Did not give him a finger, let her identify, and then one day, my child pointed to the moon with a picture of a banana, of course, she was wrong, then I told her not to say this is not the moon, this is a banana, Because she was too young I couldn't tell her the difference between the moon and the banana. I couldn't even think of how to get her to understand it. But what surprised me was that after this time she never denied missing the moon and bananas. I feel very surprised by this, whether it is real or in pictures, and I have found that our ability to learn and correct mistakes based on small samples is very powerful. This point is far beyond our current artificial intelligence.
The second story was some conversations I had with my children the other night. It was known to those who had had children. When I was four years old, I had to go to bed every night. She wanted to keep listening to music all the time. And then she told me that she said that you can never stop music. I just stole the concept, and I said I would release it again this time. It may sound a bit like an adult bullying a kid, but my kid immediately heard me steal concept and she and I Say mother, you have to keep waiting until tomorrow, tomorrow I will not listen. When I say okay, then you can rest assured that I will sleep. I again changed the concept and said that you will not stop until you fall asleep. I think that when I fall asleep, I will stop again.
As a result, I later discovered that the children had been analyzing my words. After 1 minute, she said, "I don't want to leave after I'm asleep. I'll wait till tomorrow. I say, OK, but I'm thinking, wait, After 12 o'clock, it's tomorrow. After 12 o'clock it's closed again. I think there should be no problem with it. And these are all I think. I didn't say it, but I didn't expect another minute. The little kid added: Tomorrow is, I wake up and look at the morning light. It is tomorrow. Then you can stop. At that time, I was thinking about how bright the children of God are now.
Then I thought about this problem. It is actually very difficult. If you hand it over to a computer, it may hang in the second step of the first step. And further, the child will steal the concept with me. Let me give you an example. She used my bag to play. I would tell her if my mother told you that someone else’s bag cannot be turned over? She will rightly say that mother is not someone else! What implications do these issues have for me? It's about complex semantic understanding and logical reasoning , which is a very difficult problem for our artificial intelligence, but for a three-year-old child, he can handle it very well.
The third problem is my child. He often learns by talking to himself. In fact, we don't have much time to teach him. Sometimes I will rush to say: COMEONBABY, we hurry up, there is no time. if. One day she discovered that her little flower sprouted. She asked me to see it. I grind a bit, and she came and took my hand and said, COMEONBABY, follow me! I was surprised because I never told her what this sentence meant but she could learn from context. Another point is that I often fall into a variety of bizarre questions: For a period of time, she will ask a variety of questions. For example, can mom eat this food? Can that eat? Will he later ask if he is an animal? Does it bite? You see, she is constructing his own knowledge base and knowledge map on his own initiative. This knowledge map was not told to us by people after they were built, but she built it up by herself, and this ability is still gradually developing. Rich, so this actually tells us that in human learning there is such unsupervised learning and active learning ability , she will take the initiative to ask questions, and then according to the answer to their own knowledge for further understanding and sublimation, this is our present The artificial intelligence is still far away.
The last question I will briefly talk about. One day we see a picture book of music. (above) This picture may be too abstract for us adults. This is Beethoven, when we were listening to a piece of music and looking at this picture book. She looked at the first sentence of this pattern, which means that he was not happy with her mother. I looked at it and wanted to say that he was not unhappy? But fortunately I didn't say it because I looked at the book. I found that the next sentence in the book was: We rarely find a portrait of Beethoven's smile. He always looks very depressed and angry. I am very surprised at the child's sensitivity and his understanding of emotions. This is also a very big place where I feel that artificial intelligence and the human intelligence possessed by a very small child are different: it is the perception of emotions and emotions, and the stimulation of nature . Of course, this exploration has not stopped. For example, Ma Wenchengji, he wrote some books and discussed how he views the relationship between emotion and thinking. He thinks that the two are common.
So to sum up, you see that we have so many problems. Nowadays, there are many problems that are being studied in the field of artificial intelligence, and there are still some issues that have not yet been studied. And these problems may actually be something that we humans can grasp for a two-year-old or three-year-old child. So artificial intelligence is far from being very good, so summing up my view of the present is such a sentence. This is the best time. Because I think the best thing does not mean that this thing has been very brilliant. It has come to a harvest time and there is already something very mature. That is not the best time. Because such an era has no chance for young people, the best era means that we have so much to study, we have already seen the direction, there are many directions and not only one, and then There is a constant voice reminding us that to calm down here there are bubbles that are not too lively, beware that there will be troughs, etc. and so on, so that a series of sounds will remind us. This is an era of blooming flowers. This is an era of opportunity.
So I really like the template that we think the General Assembly provided. It is a starry sky. I think since my speech started with a book, I would like to end it with a page of books. This is a poem I read to my children. I read this poem and found out that there are so many pieces of the Little Star poem. I think this is the one that best suits our theme today: we are looking for artificial intelligence. That goal is like the Twinkle Star twinkling in the sky. We are now walking in the dark, exploring the direction of the researchers, traveller in the dark. What we are looking for now is only a small part of artificial intelligence. , Tiny spark, but it's okay, For you never shut your eye, Till the sun is in the sky. Because we will never give up until the final success. thank you all.
Attachment: twinkle star
Twinkle, twinkle, little star
Twinkling stars
How I wonder what you are
What is the wonder?
Up above the world so high
Far from the world smoke cloud
Like a diamond in the sky.
Like a diamond night sky
Twinkle, twinkle, little star
Twinkling stars
How I wonder what you are
What is the wonder?
When the blazing sun is gone,
Lieyang burns off Zhou Hejing
When he nothing shines upon
The rising sun is no longer Xinghe Qing
Then you show your little light
Jingjing Lingling Hanging in the Night Sky
Twinkle, twinkle, all the night.
Flashing always does not stop
Twinkle, twinkle, little star
Twinkling stars
How I wonder what you are
What is the wonder?
Then the traveller in the dark,
Wandering in the night
Thanks you for your tiny spark,
Shallow fire with you
Could he see which way to go,
Long night long road
If you did not twinkle so.
If there is no bright spot Yingying
Twinkle, twinkle, little star
Twinkling stars
How I wonder what you are
What is the wonder?
In the dark blue sky you keep,
Deep blue night sky you figure
Often through my curtains peep,
I often see my screen
For you never shut your eye,
Never closed your eyes
Till the sun is in the sky.
Until the sun appeared again
As your bright and tiny spark
Because you are smart, light collar
Lights the traveller in the dark
Illuminate the night trip
Though I know not what you are
I still don't understand what you are
Twinkle, twinkle, little star.
Twinkling stars
Fpc Connector,Straight Pin Fpc Connector,Gold-Plated Straight Pin Fpc Connector,Conventional Fpc Connector
Shenzhen Jinyicheng Electronci Technology Co.,Ltd. , https://www.jycconnector.com