Don't dream, AI can't save your love | A live broadcast of 2042 through 2042

Sayings:

Have you ever felt the same as me?

I do n’t have much hope for love, but I occasionally want to talk about a love feelings and feel loved; I do n’t want to get married and give birth to a baby, and I am afraid of loneliness.

The work is bald and life is unsatisfactory, but I can't find someone to vomit ...

In fact, they all point to a question: "Where do people go?" We talked with international artificial intelligence experts, chairman and CEO of Innovation Workshop, and the famous science fiction writer and screenwriter Chen Yifan.

Recently, they jointly completed a book: "AI Future", studied a large number of professional information, visited a large number of AI companies, predicting how it will be in the future development set of development and how to apply in various industries. For example, if you are in love with AI, you don't have to worry about the derailment of the object; AI accompanies me to retire, at least you don't have to die alone ... but Li Kaifu has a saying a word of shock: "Don't be deceived by AI, it definitely does not really love you sincerely "".

我们还聊了很多：“可以用AI 复制去世亲人的记忆吗？”、“可以靠AI 养老吗？”、“ AI 能治好我们的心理疾病吗？”……如果你错过了直播，可以点击Let's take a look at the video below.

Zhang Wei:

Welcome to the Live live room of the New Shi Xiang.

Today's theme is "Go to 2042". What possibilities do it have after 20 years?

In the AI ​​world, the "excellent index" of convenient life and hardware will become very high, but everyone wants to know most--

If my love with people really does not have hope or encounter problems, can I solve love problems from hardware, including software, are you optimistic now?

Chen Yanfan:

A friend made a non -fiction survey of anthropology, a product called replica, which is a bit like an AI assistant to learn language. There is a virtual character to communicate with you. Just talk to AI.

After the investigation, a lot of Chinese girls used this technology platform to fall in love, far exceeding other countries. It reflects a lot of problems behind this, including the love in real life, is it very disappointing? Especially disappoint women. They may get an emotional satisfaction in a very smooth communication process with the virtual AI image and get emotional value. And this emotional value may be in real human beings, no matter men or women, TA cannot reach it.

On the one hand, the level of technology continuously improves and the level of natural language understanding has continued to improve; on the other hand, AI may know how to please ourselves than real humans, because it can analyze your preferences, analyze your subtle emotional changes when speaking Essence

I think this thing is very interesting, that is, when we talk about a relationship, what are your expectations for him?

Do you want to have a perfect object that makes you particularly comfortable and satisfy all your expectations? Or can we accept some pain, friction, and obstacles to complete self -growth?

I think this is completely different, especially the understanding of the relationship.

Kai-fu Lee:

When you fall in love with AI, I will suggest everyone:

If you see a very cute opposite sex in a Chat Room one day, you know that TA is AI, but you can't help but fall in love with TA. TA looks like you love you very much.

You must remember that TA definitely does not really love you. TA just simulates a way to love you. Don't be deceived by such AI.

The second suggestion is that many people may want to make AI as smart as people someday, and can make Super Intelligence. It's (laughs).

Zhang Wei:

Do you want to fall in love with AI or experience this relationship? Does anyone ask you if you will chat with AI?

Chen Yanfan:

I often chat with AI, and I often ask questions that are not easy to talk to ordinary people in reality. For example, do you believe in the existence of God? what is the meaning of life? I really want to know what AI's answer is. We are quite like a tree hole.

Zhang Wei:

If we are not very cutting -edge researchers and want to chat with AI, where can we chat?

Chen Yanfan:

Everyone who likes Microsoft Xiaobing must know more about this kind of Chatbot, which is basically functional in China.

For example, if you use some smart appliances, it will have an AI assistant. Then like my parents, it was particularly funny, because many household appliances in the family were smart, and the TV had its own name, and the air conditioner was also wrong.

I want to call the name TV, but it is called an air conditioner. As a result, the air conditioner is turned on, and the TV is not turned on. They have to take a small book and write down all the names of various home appliances. I think this is a particularly interesting phenomenon in the intelligent era.

Zhang Wei:

A audience asked, if your lover dies, would you be willing to give your lover's memory to AI? Do you think you will become a big trend to promote the development of technology?

Chen Yanfan:

In fact, several cases have happened recently. Domestic and foreign countries use AI to learn some of the corpus of their loved ones, and then turn it into a loved ones who are dead to chat with you, or communicate like leaving their lover.

This represents a very common human emotional needs. We will sacrifice the ancestors, loved ones, loved ones, etc., I hope to leave part of his part of us. I think this is understandable. Essence

But you imagine that in the future, if a child starts from birth, we save all his data and then hand it over to AI, then can we build a very close to the child’s chat, behavior, and reaction of this child. Model or virtual person.

Whether this virtual person also has a so -called personality, his life may be infinite.

If you put him on the server in the cloud, what we want to ask is: Is such a feeling, is it true and effective? Whether it makes you feel that I have betrayed the real loved ones and loved ones, and I pin emotions on the affiliated objects of technology.

In the end, you will find that the only real thing is that the environmental technology we build itself lies in your feelings.

For a kind of emotional connection and bond established between others and you, it evokes you a lot of memories. If you get along with this person, you must be real. Feel such a fluctuation.

So if this perspective, I think AI it is very useful.

However, it may be limited to a reasonable range. You cannot because of AI services, which causes you to feel that the true soul will be resurrected after this AI, which may exceed the boundaries of technical ethics.

Zhang Wei:

Is there any science fiction we have seen in 2022, or the science fiction we have talked about has been realized?

Chen Yanfan:

Many of Mr. Ye Yonglie's "Little Spirit Tour the Future" has become a reality. Including the cultivation of artificial meat in the laboratory, transforming crops with genetic technology, robotics of housework, and driving cars on the road, and so on. The current gene editing technology, including quantum communication, brain science, material science, space technology, many of them are in the original science fiction novels, many years ago have been imagined.

We now imagine our life after 20 years, just like we look back at 20 years ago. For example, at that time in 2002, can you predict everything we are now?

Including smartphones, most of work, entertainment, and social networking are carried out in a virtualized environment. Our travel high depends on a new identification system such as scanning code.

Therefore, I think that many technologies may not have a day -to -earth change, but its implementation and application of the specific level of society may be far beyond our expected level.

Zhang Wei:

In the past few years, have any things have made us excitedly say, wow, this has really entered the AI ​​era?

Kai-fu Lee:

I think the biggest example is Alpha Fold 2, a frontier artificial intelligence enterprise, jointly founded by DeepMind by DeepMind (artificial intelligence programmers and neuroscience Dimis Bisbis, etc.).

Because the folding of protein has always been a huge unsolved problem in the biological field. Once folded, we may be more likely to invented new drugs, which has a particularly great impact on the future pharmaceuticals of the new virus.

In the past two years, AI has been able to fold almost all human protein known, which is a great contribution to science.

On the one hand, AI solves the technical problems of humans that cannot solve it. It does not need to use any physical media at all. As long as a structure is infused, it can calculate the result; on the other hand, AI brings great benefits to future pharmaceuticals.

Chen Yanfan:

I am more interested in pre-training models like GPT-3. It understands and exchanges natural language at a very amazing level, answering many questions, and even imitating some dead celebrity scientists to communicate with you. Some creations.

For example, Zhang Wei, I remember that you were poetry lovers in college. Now AI can write some very decent verses, and even add to the creation of novels. This has brought us a great impact.

Because in the past, we felt that the work involving creativity may be the unique ability of human beings, but now some of AI's progress will make us start doubting this.

I believe that for a long time in the future, people and machines will be a long -term collaboration process.

For example, the artist began to use AI tools to create, and then there will be many traditional auction houses, which will allow AI works to participate in auction.

It entered a discourse system to gain its own value, and it will become more and more intense in the future.

Zhang Wei:

According to such calculations and derivatives, for example, we said that in 20 years, what are the possibilities?

Kai-fu Lee:

The first one may be that AI+Medical will make us live healthily and longer.

Twenty years later, we predict that people who were born at that time could live 100 years old, and they lived to 100 years old, or even lived to 120 years old.

The second may be that we can be liberated from repetitive labor.

Each of us does a lot of repetitive labor every day, such as washing dishes and scanning the floor or mechanically returning to some emails. Many white -collar "copy and paste" assistant work will be replaced.

It is not to say that a person's work is gone, but that each of us is more or less to do repetitive work. In the future Or what you like, this is also a particularly huge change.

Zhang Wei:

Regarding this change of changes, we don't want to say that waiting for you to work hard, we enjoy the results, what methods can be involved in it, do you know that there are suggestions for the teacher?

Kai-fu Lee:

Many open courses can go online. American websites include Coursera and UdaCity. You can go to Wu Enda's course on Coursera or Sebastian Thrun on UdaCity. They are all AI masters, and they are also very easy to speak.

In addition, I think you can buy some popular science books, including the books I wrote before, you can learn simple AI programming, and use Python to do some simple AI experiments.

In fact, this technology is too important in the future. Understanding some fur can help us know when AI is useful to help our career development.

Learning AI technology can also cultivate our logical thinking, but don't think that everyone must become an AI expert.

Just like everyone is the beneficiary of computer software and mobile phones today, but it does not mean that we will all make a software or PC and mobile phones.

Zhang Wei:

If I am difficult to determine whether it is duplicated labor, and then, if my work can be replaced, except for the service industry, can I think about what I will do in the future? Do you have to prepare for future work? In other words, I want to completely abandon my current skills and get a place in the future.

Kai-fu Lee:

This job will not come so fast, and it will not say that all jobs will disappear within one minute.

For example, as counter staff in the bank, after online payment and mobile payment, the counter did not disappear, and they can still do more advanced, difficult, and more communication.

We can use this as a case, that is, in your work, you can imagine that it is not the entire work, but that some tasks are replaced. So which are more difficult to replace? Do you have to spend time to do this better?

In addition, some work will indeed be completely replaced, and the practitioners of these tasks must re -train new skills as soon as possible.

Of course, it is best to use past experience, but it is often necessary to re -train things that AI cannot do. It is usually not particularly mechanized and special. It may require some reasoning thinking, communication, gathering knowledge in different fields, etc. Essence

There are some jobs, you will not take it as a service industry, because the proportion of communication between people is too large, and AI cannot be competent for a while.

For example, sales and markets are relatively difficult to disappear, and the repetitive work of simple finance, legal affairs, and text is relatively easy to be replaced.

In this book, I have a two -dimensional picture. You can see the speed and dimension of work disappearing in the picture. Everyone can make some evaluations.

Finally, think about whether every decision needs to think more than 5 seconds in my work every day.

If most of them are not needed, it is repetitive work; if necessary, the difficulty and probability of replacement will be reduced.

Zhang Wei:

We received a lot of users.@Mmg asked AI pensions might be achieved or even popularized after 20 years? Another user@另 另 is also asking if AI can take on the heavy responsibility of accompanying the elderly?

Everyone may have an uncertainty or insecure in the future of the elderly. Many people may feel that I do n’t want to have children. I want to be alone. Then I may think more about this.

What can we live in the elderly in the future of AI for medical care, especially at the elderly party?

Kai-fu Lee:

In theory, AI can do a lot of care for the elderly: helping the elderly cook for dishes, accompany the elderly to help them take a bath, etc., can achieve 20 years or even a long time.

But we still have to think about the structure of the entire social structure. If many repetitive work disappear, what should people do?

I think the essence of the first issue needs to be concerned. Many current blue -collar and white -collar workers will disappear. The remaining work has many creative invention strategies. CEOs, artists, and scientists have relatively high thresholds.

But what new job opportunities will they have now?

I think the service industry will still exist, and it will require more people. The service industry has temperature, and the communication between people's communication between people can always do better than machines.

It is difficult to imitate the delicate emotional care of people. Even if it is well imitated, I think most people cannot accept the cold machine to "pretend to be human care" themselves.

But the machine has no heart and has no real thinking, but is just cared for you by Program. I think most people will not accept it, especially the elderly need a loved one or at least the care of people, not the care of the machine. The second question is that the old man really wants the robot to help him take a bath and chat, or do you want someone to help him do this?

We actually saw a entrepreneurial project, and we did robots to accompany the elderly. Finally, we found that the old man's favorite function is customer service.

The old man does not really need customer service help, but wants to find someone to chat with him. Why do n’t my child come to see me today? You can see how cute my granddaughter is, and the old man desires people's companionship.

In the future, practitioners who drive, wash dishes, etc. are facing the possibility of being replaced by machines. In fact, you can consider transforming care -oriented work like the elderly. This is also the justaries in the future society.

On the one hand, laid -offers have channels to find new jobs and the next stage of life.

So I think that in the future, I hope there will be many intelligent things to help the elderly prevent falling.

Remind the elderly to take medicine, remind the elderly to see the doctor, and find out the symptoms of the elderly, which helps the elderly's safety and health to a large extent.

But it is a thing hidden in the house. It is not a robot jumping out and pretending to be your serviceman or companion.

We should still let people do these services and warm care work.

Chen Yanfan:

By communicating with AI or many robots, you can find something:

People are constantly projecting some of their expectations on individual other objects.

We will always be self -centered, including you like a kitten and puppy, then you will naturally understand many of its behaviors and understand that it has emotional dependence on you and has emotional interaction.

So I think this kind of thing will also happen in an interaction between people and machines and AI. Many times not say that the machine has any emotions to you, but you want to feel this temperature from him.

I think the younger generation is very used to this kind of two -dimensional, the love of paper people, virtual idols, etc., that is, everyone has accepted such a new type of emotional relationship.

But I also believe that Teacher Kaifu said--

In the end, although all this can be done perfectly, like people are close to people, you still feel that the real connection and exchanges between people are still unable to be replaced.

Zhang Wei:

With the development of AI, such as depression and anxiety, will problems change because people love alone? Is AI's help for the treatment of mental health diseases?

Kai-fu Lee:

I think there are some signals that make us more optimistic.

In the United States, it is not realistic whether a psychiatrist or a psychological counselor is expensive to accompany them at any time.

For most people, they cannot talk to a psychologist every day, but AI can be there at any time. It is effective and complementary to do a certain companion function.

Human doctors cannot be completely replaced by AI, but they can be a "AI Assistant" that responded at any time at any time. We see someone using AI to make some hotlines to help those who may temporarily play this hotline.

If some autistic children can accompany them with AI, this should be done very carefully, because they don't like to contact people, how to do such an environment is still a scientific research topic.

It may be used to use the method of VR and the method of the universe, but it can't make them feel too realistic, so I think this should not be completely resolved, but I hope to use AI for subsidies.

Just like we can see AI as a doctor and teacher assistant, we should also be psychological counseling, and so on.

Zhang Wei:

We now give many repetitive labor the value of happiness, such as washing dishes may be part of life, and so on.

When we look at it objectively, when I really do not need to wash the dishes, people may have more opportunities to find or re -discuss what happiness is.

Because of the arrival of the AI ​​era, there is no labor, will happiness become easier or more difficult?

Chen Yanfan:

I think different times, everyone's definition of happiness is very different.

For example, my parents grew up in the age of lacking supplies. A little material satisfaction, they would feel very happy. By the time when we were born, material was relatively rich.

So after 20 years, according to your location, the work they are engaged in (happiness) will be very different, and the definition of labor is different.

It may take more time to engage in work that is more creative, more mechanically repeated, or has more contributions to the entire society to bring us spiritual happiness and satisfaction.

I believe that in that era we used a kind of contribution to the sustainable development of the entire society, and using such a good standard as a source of happiness.

Maybe this answer is more positive, but I really think so.

Kai-fu Lee:

I think people's happiness is actually cultivated in these thousands of years, including our DNA, self -adaptation difficulty, etc. This adjustment takes time, but AI is indeed very fierce, so we really need to go very much Do some adjustments.

Today, it may be based on short -term sensory stimuli. It can see very refreshing content and earn more money. These things really need to be eliminated. Such things will change greatly. win.

Also, self -actualization will become more and more important. The significance we exist, the value we can create, how can we maximize our contribution and influence on the world?

I hope that there are various new methods, people have greater wake -up and awareness, which is a large extent a human evolution.

At the end

Back to our original question: "Where should a person go?"

Maybe you can find your own answer from the book "AI Future". From a technical perspective, they have launched what level the AI ​​will reach in 20 years in the most logical way. Everything that may happen in the future will be rehearsed in front of us.

But in any case, I want to tell you: "We will find a way, we always have a way."

Edit: Yuan Sanjian

- END -