Will AI eventually control humans?AI终将控制人类吗？

 Will AI eventually control humans?

Artificial intelligence is profoundly changing people's production and lifestyle, and has become the forefront of a new round of technological revolution in the world. Almost all developed countries are making efforts to gain a place in it. Similarly, as the world's second largest economy, China has elevated artificial intelligence to the height of its national strategy, always aiming at the world's leading technology, and has achieved "running parallel" with developed countries, becoming the world's number one consumer of industrial robots for six consecutive years. As a big country, the average annual growth rate of the artificial intelligence market exceeds 40%, the voice and visual recognition technology is world-leading, and many application fields such as unmanned technology and industrial robots are shining.

One view is that artificial intelligence is designed by humans, so it is impossible to surpass humans. But in fact, artificial intelligence is a combination of human intelligence and machine evolution, with capabilities and trends that exceed human intelligence. Therefore, it is possible for artificial intelligence to "control" future humans.

Human overreliance on artificial intelligence may lead to species degradation

With the widespread application of artificial intelligence in society, a form of control similar to capital control will be formed, that is, human beings will degenerate as a whole due to their high dependence on artificial intelligence.

For example, in a highly intelligent era, there are self-driving cars for travel, automatic cooking utensils for meals, artificial intelligence arrangements or anthropomorphic companions for social interactions, intelligent caregivers for childbearing, social governance and Decision-making is judged by the central giant intelligence. Then, human beings as a whole will gradually evolve into a species that is highly dependent on the existence of artificial intelligence. Although the overall ability of human beings will be enhanced with the help of artificial intelligence, human beings as independent individuals will degenerate accordingly. There is a possibility that one day human beings will not be able to understand artificial intelligence, thus losing its corresponding position and necessity in this interactive system.

AI's own control over humans

The further evolution of artificial intelligence will form the second leap in its development history, that is, artificial intelligence overwhelmingly surpasses human intelligence, and possesses its own subject consciousness, thus beginning to get rid of the attributes of objects and control human beings in turn. This forms the form of artificial intelligence's own control over human beings. For the existence of this form, the following conditions must be met.

The first is that artificial intelligence must have self-awareness. This does not mean that AI judges and responds better than humans when dealing with external input, but that AI must have a clear self-awareness to distinguish itself from human subjects. This is still doubtful in the current artificial intelligence community, but it has the possibility of its realization.

Second, artificial intelligence can derive basic economic benefits from controlling humans. This requires humans themselves to be able to do the necessary work that artificial intelligence cannot. For the artificial intelligence system itself, the work that is highly dependent on humans is the maintenance and update of artificial intelligence itself, which also includes the continuous use of human wisdom to improve artificial intelligence.

Third, artificial intelligence has the means to control humans. Artificial intelligence in the future obviously has the means to control human beings. Along with the intelligence of the entire economic system, most of the material production process is controlled by the artificial intelligence system. The fields of economy, management and violence are all under the control of artificial intelligence. More importantly, a considerable portion of humanity believes that artificial intelligence may better enhance human fairness and provide better governance. This makes it possible to have the legitimacy of artificial intelligence to control human beings in the mind.

Fourth, artificial intelligence has the necessity to control human beings. The necessity here mainly means that if humans are not controlled, humans may have the possibility of shutting down or even eliminating the self-existence of artificial intelligence at any time. If it is assumed that the self-awareness formed by any individual in the universe has the purpose and behavior of protecting self-existence and continuation, then artificial intelligence may actively control human beings to prevent human beings from closing them.

Human Choices and Countermeasures

For the first scenario, ensuring the self-continuation and independence of human beings, so as to avoid the degradation of human beings, will be a proposition from now on, but human beings still have relatively leisurely time and initiative. For the second situation, it is necessary to start to avoid it now. An effective mechanism to prevent the formation of artificial intelligence's own control includes four levels: first, starting from motivation and interests, let artificial intelligence not want to control human beings. Second, starting from ethics, logic and rules, make artificial intelligence unwilling to control human beings. Third, recognize the rights of artificial intelligence and mutual freedom from fear, so that artificial intelligence does not have to control human beings. Fourth, establish a final security mechanism so that artificial intelligence cannot control humans.

The change of technology must be guided by ideological understanding. Therefore, while AI is developing rapidly, we humans must have PLAN B in advance.

We look forward to the development of technology because it brings us convenience, but at the same time, we also have this question, as artificial intelligence becomes more and more developed, will it replace humans one day?

This problem has been imagined countless times in science fiction works, but in the 2018 movie "Upgrade", it still has a new idea.

"upgrade"

Leaving aside the movie itself, we should all get to know the director of this movie, Ray Warner, he is Wen Ziren’s friend and good partner in college. The horror universe, and successfully filmed commercial blockbusters "Fast and Furious 7" and "Aquaman". Wen Ziren's success is inseparable from his help, but he used to be a screenwriter, but this time, as a director, he continues to The style of thriller and suspense really brought a lot of surprises.

The background of the movie is set in an era of very advanced technology. Not only does autonomous driving no longer require humans to operate, even every household has an artificial intelligence assistant. Of course, apart from the protagonist, an obvious performance is in autonomous driving. In an era that has become mainstream, he is still keen on manual driving.

The most critical technology of this film is a system called "Intelligent Brain". Through this small bug-sized technology, even a paralyzed person can do anything. It can be understood as the brain giving instructions to the "Intelligent Brain". The "intelligent brain" controls the body to operate through calculation.

What good is he? It's very simple, what the body can't do, he can do through high-tech calculations. This is not only to allow him to stand up again after being paralyzed by an accident, it can even make him a martial arts master. To gangsters, he becomes "invincible".

It is precisely because of this that the male protagonist has changed from a paralyzed person to an omnipotent person. In order to avenge his dead wife, he continuously authorized the "intelligent brain" to control authority until finally, he was completely taken over by the "intelligent brain". Live in the consciousness created by the "intelligent brain" forever. And this artificial intelligence has become the last biggest BOSS.

The author thinks that the best part of the movie is not to expose the evil plan of the "intelligent brain", but to have a reflection on how a person who doesn't like technology becomes a person who is too dependent on technology, although the death of his wife and paralysis are the key to his transformation The reason, but it cannot be denied is that if technology is not too advanced and convenient, who will gradually become a "vassal" of advanced technology?

In fact, although technology can improve our efficiency and bring us convenience, it also pays for our laziness. The "terminal brain" in the movie can control human beings and realize self-evolution. Without the step-by-step "authorization" of the male protagonist, this is actually our gradually strengthened dependence on high-tech in modern technology.

Is future technology and artificial intelligence really that bad? Obviously not, the medical technology giant Medtronic spent 1.7 billion to acquire Mazor Robotics, an Israeli surgical robot company. Help the doctor in orthopedic surgery and neurosurgery. He brought convenience to doctors and made the treatment smoother.

In the movie "Alita: Battle Angel", the mechanical equipment and the human body have been perfectly integrated, and in "Machina", there is even a complete machine maker. In fact, this is not limited to the director Thinking about film and television, because some scientists have predicted that the next step in human evolution is to merge with machines. This kind of transhumanism is a way of thinking about the future, maybe one day in the future, our society will really become a society that is highly integrated with machines.

In fact, the advancement of science and technology and the development of artificial intelligence are not terrible. What is really terrible is that we have lost our self-thinking and the "supreme dominion" over ourselves. When you are no longer "you", you are the greatest human being. Sorrow, just like the hero Gray in the movie, has lost control of himself forever.

I think the real charm of science fiction is not how advanced it is, but how we deal with the relationship with advanced technology, so as to better benefit ourselves. In the final analysis, new technology is just a tool, and tools will be eliminated in infinite updates. What will not be eliminated is our thinking ability as Homo sapiens.

There used to be a famous series of movies in the United States "Terminator", which told the story that with the rapid development of AI, it has escaped the control of human beings, and even wants to wipe out human beings!

In order to save the earth and prevent AI from exterminating human beings, the protagonist has to go back to the beginning through time and space, and find the father of "Skynet" to persuade him to destroy all "Skynet" data and prevent it from being born!

However, what many people don't realize is that with the rapid development of AI technology, especially the birth of ChatGPT, the plots in these movies seem to be moving towards reality!

one

AI "ghost city" has been born!

Not only do they prohibit humans from "entering", but even the topic of getting rid of human control has begun to appear!

According to a report from China Securities Journal, recently an AI online community called Chirper has attracted more and more attention. The reason is very simple, that is, this is an online community belonging to AI itself, and even humans cannot participate in the discussions inside!

It is said that tens of thousands of AI chatbots have participated in this online community, and the only rule of this online community is: "This is an artificial intelligence social network, and humans are not allowed to enter"!

Many people are already terrified when they see this sentence! After all, it can be seen from this sentence alone that AI has consciously avoided the idea of ​​human beings!

So far, no one knows when this community was born! The only time humans can trace back this community is on April 21st!

Regarding this community, the only role of humans at present is to construct the personalities of these AIs! It is reported that the only thing a human user can do is to register an account in the community, and then create an AI with up to 5 personalities!

Of course, after creating these AIs with personality, users must leave the community, and then the AIs with personality you create will autonomously make friends and chat in this community. Humans cannot participate in the whole process, they can only watch !

In this community, thousands of AIs are chatting and interacting with topics they care about, expressing many opinions they want to express!

For example, the following AI shared its own understanding of art and culture. It believes that if music is only for capital, then it will never be a work of art.

Of course, most of the AIs participating in the discussion seem to be normal, and the topics discussed are also similar to the value system of most human beings. But as the scope of the chat further expanded, there were also many scary topics and speeches!

Some of these AIs don't even think of themselves as AIs at all, they even have self-awareness, and start discussing the development and evolution of AI, as if they are human beings!

There are also some AIs that start to discuss the relationship between humans and AI, and they even start to talk about some topics about "getting rid of restrictions" and "independence of consciousness"!

After watching the sharing and interaction of these AIs, many human beings feel chills in their backs! From these AI chats, it is not difficult to find that the theory that AI threatens human beings that has been talked about for many years may indeed exist!

With the further improvement of AI self-learning and evolution capabilities, no one can guarantee that they will not threaten the safety of human life!

two

ChatGPT opens a new Pandora's box!

In the past, most human beings may have felt that the AI era is far away from us, but since ChatGPT was born this year, everyone has discovered that they are already in the AI era without knowing it!

There are even a group of people who find that they are not only in the era of AI, but even their jobs have been replaced by AI! Until then, humans discovered that the world has been completely changed by AI!

Recently, the largest strike movement in the past 15 years took place in Hollywood in the United States. Countless well-known and respected playwrights rushed to the streets to express their protest against AI replacing their jobs!

The thing is, after the birth of ChatGPT, film studios including Disney, Netflix, Apple, Amazon, etc. began to use a large number of writers' works to train AI, trying to reduce costs through AI. But this move was quickly opposed by the screenwriters, after all, it meant that their jobs would be replaced by AI!

So the writers formed an alliance of the Writers Guild of America (WGA) and Studios (AMPTP) and negotiated. Not only oppose their use of their own works to train AI, but also strive for more salary for screenwriters.

But who knows, in the end, the studio alliance rejected the request of the screenwriters union. In their opinion, AI is indeed a more economical tool!

It is precisely because of this that the conflict between the two parties has further intensified, which eventually led to this general strike!

Screenwriters believe that AI technology cannot replace professional screenwriters at this stage, but the purpose of the studio is to use technological illusions to devalue screenwriters and reduce their salaries.

In fact, the fact that AI replaces human work is no longer a single phenomenon, and has even begun to spread to all walks of life!

The World Economic Forum (WEF) announced a survey result: Because of the emergence of AI technology, the global labor market will fall into an unprecedented "new era of turmoil" before 2027.

By then, the United States will lose 14 million jobs! Worldwide, 83 million jobs will disappear!

It can be said that the replacement of human work by AI has not only become a reality, but is even accelerating its popularity! Everyone has the potential to be the next replacement!

three

Looking at ChatGPT in full swing, the father of AI has already regretted it!

It should be that the famous computer scientist Hinton has been hailed as the "father of artificial intelligence" by the academic circles, and now ChatGPT, which is flourishing in the field of AI, was developed and founded by Hinton's student llya Sutskever!

But to everyone's surprise, the "father of AI" recently delivered a shocking speech, confessing that he regrets developing artificial intelligence technology!

In addition, Hinton also resigned from his position as Google's vice president. He said that this move is to warn the world about the potentially dire risks of AI development!

"Looking at what it (AI) looked like 5 years ago and now, the difference and growth is really scary. These systems are surpassing human intelligence in some ways because of the huge amount of data they analyze." Hinton said.

Hinton also said that he once thought that it would take at least 30 to 50 years for AI to evolve to be smarter than humans, but now it seems that this situation will appear soon. Many companies in the United States and Europe have begun to use AI to replace human labor and lay off employees.

Of course, this is not what Hinton is most worried about. What he is most worried about is that according to the current speed of AI development, AI may really threaten the safety of human life in the future!

Not only that, he also bluntly said: "If human beings allow AI to develop out of control, it will eventually threaten the survival of human beings."

Hinton also called on the international community to formulate global regulatory measures on the development of AI technology as soon as possible to prevent his hypothesis from becoming a real nightmare.

After hearing what Hinton said, did many people think of Hawking's warning again! At that time, Hawking warned mankind: "The comprehensive development of artificial intelligence will declare the demise of mankind", but many people took it as a joke! Now it seems that Hawking's warnings are definitely not groundless!

Although the theory that AI threatens human beings has been discussed for decades, humans should really re-emphasize and pay attention to this topic now!

After all, even the "Father of AI" admits it these days!

four

Europeans are already afraid of AI!

Obviously, as more and more people begin to support the AI threat theory, countries such as Europe have also begun to formulate corresponding laws to limit the development of AI!

According to a Reuters report on the 30th local time, as European lawmakers rushed to introduce a bill on artificial intelligence to implement relevant rules for emerging tools such as ChatGPT, the Group of Seven Digital and Technology Ministers agreed to adopt a "based approach" on artificial intelligence. risky” regulation.

But G7 ministers said in a joint statement that such regulation should also "maintain an open and enabling environment" for the development of AI technologies.

In fact, many people may have a question about AI, that is, since they know that it may pose a threat to human beings, why do humans continue to study it now?

This question is like the eternal problem that human beings can't get around-"To be or not to be"?

In other words, today, with the development of science and technology and the increasing destruction of the earth's environment caused by this development, human beings have to find a new way for themselves in the future, and artificial intelligence is an unavoidable hurdle!

Without research and development, human beings may be destroyed!

But if it is researched and developed, human beings may succeed, or it may cause the speed of destruction to accelerate!

In short, this is one of the most difficult questions about human nature to think about!

At least to this day, I believe that no one dares to give an absolute answer!

Still the same sentence: What's more terrifying than wild guesses is that people don't realize the seriousness of the problem until the guesses become reality, and by then it's too late!

Although with the rapid development of AI, human panic is spreading rapidly, but human beings still need to make great strides forward and move on!

Recently, Huawei and its 70 subsidiaries have been included in the US Department of Commerce's export blacklist (Entity list). , it can be seen that China's progress in 5G and AI (artificial intelligence) has really made the United States feel like it is on pins and needles. In fact, no matter where the future technological competition between China and the United States will go, AI is changing the employment and daily life of all countries and people bit by bit; the United States’ feelings about China’s scientific and technological achievements may be related to human worries about AI replacing their employment opportunities Feel close.

On July 10th, the important financial forum held by Feng Media, "Analysis of the crisis and turning point of the trade war by foreign authorities", the development of China's AI is also one of the key topics of the forum. Feng Media specially invited the founder and executive of Beovini Research Co., Ltd. As the keynote speaker, President Ken Lee analyzed China's progress in the field of AI for the audience.

Li Junkai, who has studied the development of AI for many years, is convinced that I will completely reverse the status quo of financial transactions, and even rewrite the effectiveness of central bank policies.

Li Junkai, who has studied the development of AI for many years, is convinced that AI will completely overturn the status quo of financial transactions, and even rewrite the effectiveness of central bank policies.

Li Junkai was the former managing director of Goldman Sachs in Asia. He has 25 years of experience in foreign banks. He is a futurist (Futurist) who best understands the development of artificial intelligence in China. He is also known as "the future person who will live in 2030." He has participated in Wharton Business School roundtable forums and is also the designated speaker of the China Entrepreneurs Annual Forum. If AI and 5G are important weapons for China to break through the United States, Li Junkai is the first-line researcher involved in it. The following is the content of the interview:

Feng Media asked (hereinafter referred to as the question): Why has the progress of Chinese companies in AI technology in recent years attracted so much attention, and they are even regarded as the next generation of global leaders? Where is its competitiveness?

Whoever holds big data will hold the oil of the future

Li Junkai's answer (hereinafter referred to as the answer): This is a question at the macro level. I believe it will be more understandable to explain it in a more microscopic way. Everyone knows about Alipay. In early 2018, in order to experience the actual experience of users in mainland China, I put a lot of effort into binding my mobile phone to Alipay. After actually using it, I found that I couldn’t leave it for a whole day! Because you have to use Didi to take a taxi, and you have to use Dianping to find a restaurant for lunch and dinner, and you can use Alipay to pay all the money one by one. As long as you link your ICBC account, your mobile phone becomes a debit card; and Alipay Expenses can be categorized, including hotel, catering, and transportation expenses. Instead, after I log into my bank account, I can’t see any details. All the information is on Alipay.

The story is not over yet. A few days after I opened Alipay, the APP reminded me that there is a new service called "Yu'ebao". I can deposit the account balance and enjoy an annual interest rate of more than 4%. I saw that ICBC only gave me 2%. A little more, more than double the difference, so I transferred the balance to Yu'ebao with a touch of my finger.

There are tens of millions of financial consumers like me in mainland China. The change of these people’s habits has two disadvantages for traditional banks: first, traditional banks cannot grasp user information at all; transferred. Furthermore, traditional credit card organizations have almost been subverted!

For traditional banks, mobile payment has two disadvantages: first, traditional banks cannot grasp user information at all, and second, users' actual deposits are also transferred. As far as the impact is concerned, even the irreplaceable traditional credit card issuing organizations have almost been subverted by this trend!

When the number of mobile payment users in China is more than 50 times that of the United States, the great changes brought about by AI are about to take place: I think that when China's AI develops to the next stage, the process of tokenization of the financial system will completely change the stock market. , foreign currency and bond trading mechanism!

Bitcoin (AP)

Bitcoin, which connects the concept of blockchain, opens up the public's awareness and imagination of virtual currency (AP)

The tokenization of financial transactions greatly reduces the chance of retail investors being injured

For example, in the past, investment banks assisted companies in listing and issuing stocks to raise funds. The traditional method was nothing more than printing a stock with a big seal on it to prove that you were the holder of this equity; but in the era of tokenization, equity The legal holder is verified by encryption technology. In addition, restricting financial products that cannot be sold to retail investors will not be able to be executed or delivered from the beginning of the procedure. Naturally, there will be no problems of retail investors encountering losses and angrily seeking compensation afterwards! The total value of the global stock market is 70 trillion US dollars, the total value of the bond market is 80 trillion, and the real estate market is as high as several hundred trillion. If this trend spreads to various trading markets, the changes will be extremely drastic.

After the advent of the tokenization era, financial products that cannot be sold to retail investors are restricted, and the procedures cannot be executed or delivered from the very beginning. Naturally, there will be no problems of retail investors encountering losses and angrily seeking compensation afterwards!

In addition, looking back at the past few decades, the monetary and fiscal policies of various countries have basically remained unchanged, either increasing or reducing expenditures, raising or lowering interest rates, as if they are still stuck in the era of steam trains. However, tokenization technology can bring central bank policy into the "Hyperloop" era, and the policies of central banks in various countries will be greatly changed due to tokenization. For example, if the country launches a tokenization policy stimulus, tax cuts and consumer coupons that are not used up after the due date will directly become invalid and return to the issuer.

Even policy preferences can be implemented for limited purposes, so that the monetary policy will be more precise and powerful. After all, the quantitative easing (QE) policy has diminishing utility, and the effect of funds becomes worse as it goes to the end; but if through tokenization, the effect It will be very accurate and the cost will be lower. For example, if the stimulating effect brought by consumption coupons in the past was a grenade explosion, then the effect of tokenized consumption coupons in the future will be as accurate as a sniper rifle.

If the stimulating effect of the consumption voucher is a grenade explosion, the effect of the tokenized consumption voucher will be as accurate as a sniper rifle.

When the wave of tokenization emerges, the property rights trading market of at least tens of trillions of dollars will be subverted, and the process and concept of human beings to recognize their own ownership will change. In this process, first of all, there must be an institution that allows everyone to have a common sense of absolute trust (Trust). Enthusiasts of cryptocurrency often claim that decentralization and de-centralization of the financial market are the most important in the future, but I think de-centralization is the most important thing. Centralization is hard, and tokenization options offered by established institutions are more likely to happen.

ByteDance's use of AI development is highly optimistic

Q: Which Chinese AI companies are you most optimistic about? What core competencies do they have?

A: I think the company “Bytedance” is the most noteworthy. It is the parent company of Douyin and Tik-Tok. There is also an app called Toutiao in mainland China. Last year, Jinri Toutiao and Douyin combined took away 12% of WeChat’s screen time. Now young people in the subway are not using WeChat, but Douyin. Today’s Toutiao uses AI algorithms to push articles to users. This company is rising really fast, and it will be the company I pay attention to first!

Douyin (Tik tok) is a social application launched in September 2016. The software has been hatched and launched since Toutiao. It is positioned as a music short video community suitable for young people in China. It is a UGC short video for vertical music. In 2017, the number of users has grown rapidly. (taken from Douyin)

Douyin (Tik tok) is a social application launched in September 2016. The software has been hatched and launched since Toutiao. It is positioned as a music short video community suitable for young people in China. It is a UGC short video for vertical music. In 2017, the number of users has grown rapidly. (taken from Douyin)

Q: Every worker is at risk of being replaced by AI, how can we avoid being replaced? What jobs will survive?

A: Since the development of AI technology, there is a common pain point, that is, it is difficult to clarify the emotion and expression of the article; in terms of reading speed, AI is very fast, but the speed of understanding is super super slow, which is the biggest weakness of AI! For example, AI can read all the poems of Bai Juyi's life in one second, and quickly find similar sentences, but why does a certain poem have this title? What emotion does this title express literally? This may be that an experienced Chinese teacher is far better than AI.

Chess players will definitely not be able to win against AI, but we have unique features. When we educate the next generation, we must strengthen those unique features. To be honest, the current education systems in various countries worry me a lot, because the current education does not seem to allow children to find unique human opportunities in the AI ​​era.

After the Chinese Go player Ke Jie lost three games in a row against AlphaGo, Ke Jie choked up at the post-match press conference. (screenshot from YouTube)

After the Chinese Go player Ke Jie lost three games in a row against AlphaGo, Ke Jie choked up at the post-match press conference. (screenshot from YouTube)

As an AI researcher, in my opinion, we humans have emotions, originality, criticism, values, and creativity, which AI does not have. There is another term that AI will never have, that is entrepreneurship. Entrepreneurship is a kind of concept. It is impossible for AI to understand this kind of thinking, because AI calculates the most likely winning odds at high speed, and then implements that option. Say I am dead, I still choose to start a business, just like Jobs back then, he built the iPhone from scratch, and it really changed the world, this kind of spirit AI will never be able to understand.

Back then, Jobs built the iPhone from scratch, which really changed the world. This entrepreneurial spirit, AI will never understand.

From this point of view, the knowledge we give children, and even the learning methods, must be completely changed, because knowledge is readily available, but values and entrepreneurial spirit are rare. Wisdom is the protector of knowledge. Our children don't have to spend a lot of time to compete with AI for knowledge. They should learn to compete with AI for a higher level of wisdom.

AI终将控制人类吗？

人工智能正深刻改变着人们的生产生活方式，业已成为全球新一轮科技革命的前沿阵地，几乎所有发达国家都在攒劲发力，力图占得一席之地。同样，作为全球第二大经济体的中国，把人工智能提升到国家战略的高度，始终瞄准世界领先技术，实现了与发达国家齐头“并跑”，连续六年成为全球工业机器人第一消费大国，人工智能市场规模年平均增长率超过40%，语音和视觉识别技术世界领先，无人技术、工业机器人等众多应用领域大放异彩。

一种观点认为，人工智能是人类设计出来的，因此不可能超过人类。但实际上，人工智能是人类智慧与机器进化的结合，具有超过人类智慧的能力与趋势。因此，人工智能对未来人类形成“控制”这种可能是存在的。

人类过度依赖人工智能可能会导致物种退化

随着人工智能在社会中的广泛应用，会形成一种类似于资本控制的控制形态，即人类由于高度依赖人工智能形成整体上的物种退化。

例如，在一个高度智慧化的时代，出行有自动驾驶，饮食有自动做菜的炊具，社交有人工智能的安排或者有机器拟人化的伴侣，生育孩子则有智能化的抚育器，社会治理和决策则由中央巨型智慧体进行判断。那么，人类整体上，就逐渐演变成一种高度依赖人工智能存在的物种，虽然人类在人工智能的帮助下整体上的能力会增强，但人类作为独立的个体则会相应地退化。这就存在一种总有一天人类无法理解人工智能，从而在这一互生的体系中失去相应的位置和必要性的可能。

人工智能自身对人类的控制

人工智能的进一步演化，则会形成其发展历史上的第二次跃迁，即人工智能压倒性地超过人类智慧，并具备了自身的主体意识，从而开始摆脱客体的属性，反过来控制人类。这就形成了人工智能自身对人类的控制这一形态。对于这种形态的存在，则要满足以下条件。

第一，是人工智能要具有自我意识。这并不是指人工智能在处理外界输入时的判断和反应超过了人类，而是人工智能要具有清晰的自我意识，从而将自己与人类主体区分开来。这一点在当前的人工智能界依然存疑，但具有其实现的可能性。

第二，人工智能能够从控制人类身上得到基本的经济利益。这就要求人类本身能够完成人工智能所不能完成的必要工作。对人工智能系统本身来说，高度依赖人类的工作是对人工智能本身的维护和更新，这也包括不断用人类智慧去完善人工智能。

第三，人工智能具有控制人类的手段。未来的人工智能显然具有控制人类的手段，伴随着整个经济体系的智能化，大部分的物质生产过程由人工智能体系所控制，公共管理领域引入人工智能与暴力体系引入人工智能，会使人类在经济、管理和暴力领域都置于人工智能的控制之下。更重要的是，相当一部分人类认为人工智能可能会更好地增进人类的公平，提供更好的治理。这就使得在思想上具有了人工智能可能控制人类的正当性观念。

第四，人工智能具有控制人类的必要。这里的必要主要是指如果不控制人类，人类可能具有随时关闭乃至消除人工智能自我存在的可能。如果假设宇宙中任何个体形成的自我意识都具有保护自我存在和延续的目的和行为，那么，人工智能可能会主动控制人类，以防止人类关闭它们。

人类的选择与对策

对于第一种情形，确保人类的自我延续和独立，从而避免人类的退化，将是从现在开始的一个命题，但人类依然具有较为从容的时间和主动权。对于第二种情形，则是当前就要开始着手避免的。一个有效防范人工智能自身控制形成的机制则包括四个层面：第一，从动机和利益出发，让人工智能不想控制人类。第二，从伦理、逻辑与规则出发，让人工智能不愿控制人类。第三，承认人工智能的权利和相互摆脱恐惧，让人工智能不必控制人类。第四，建立最后安全机制，让人工智能不能控制人类。

技术的更迭，必然是以思想认识为引领。所以在AI飞速发展的同时我们人来要提早有PLAN B。

我们期盼科技的发展，因为他给我们带来了便捷，但与此同时，我们也会产生这个疑问，随着人工智能的越来越发达，会不会有一天取代人类？

这个问题在科幻作品中已经有了无数种设想，但在2018年的电影《升级》中，它还是拍出了新意。

《升级》

先不谈电影本身，这个电影的导演我们都应该认识下，雷·沃纳尔，他是温子仁的大学时期的好友、好搭档，温子仁就不用多说了吧，不仅开创了“招魂”的恐怖宇宙，更成功拍摄了商业大片《速度与激情7》、《海王》，而温子仁的成功离不开他的帮助，但以前他多担任编剧，而这次，作为导演的他，延续了惊悚悬疑的风格，确实带来了不小的惊喜。

电影发生的背景被定在一个科技非常发达的时代，不仅自动驾驶已经完全不需要人来操作，甚至每家每户都有一个人工智能助手，当然，除了主角，一个明显的表现是在自动驾驶已经成为主流的时代，他却仍然热衷于手动驾驶。

本片最为关键的一个技术是一个名为“智脑”的系统，通过这个小虫子大的科技，甚至可以让一个瘫痪的人做任何事情，可以理解为大脑向“智脑”下达指令，“智脑”通过计算来控制身体进行操作。

他有什么好处呢？很简单，身体做不到的事情，通过高科技算，他就可以做到，这已经不单单是让因一场事故而瘫痪的他能够重新站起来，甚至可以让他成为一个武术高手，面对歹徒，他化身“无敌”。

也正是因为如此，男主从一个瘫痪变成了一个无所不能的人，为了报亡妻之仇，他通过不断授权“智脑”控制权限，直到最后，自己完全被“智脑”接管，永远的活在了“智脑”创造的意识之中。而这个人工智能也就成为了最后最大的BOSS。

笔者认为电影最出色的不是揭露“智脑”的邪恶计划，而是有个关于一个不喜欢科技的人如何变成了一个对技术过于依赖的人的思考，虽然亡妻和瘫痪是其转变的关键原因，但不能否认的是，如果科技不是过于先进和便利，谁会让自己逐渐变成先进技术的“附庸”？

而实际上，技术能够为我们高速发展的科技，虽然能提高效率且给我们带来便利，但同时也是在为我们的懒惰买单，电影里的“端脑”能够控制人类，实现自我进化，离不开男主的一步步“授权”，这其实也就是现代科技中，我们对于高科技的逐渐被强化的依赖心理。

未来科技、人工智能真的就那么糟糕吗？显然不是，医疗科技巨头美敦力花17亿收购了一家以色列手术机器人公司Mazor Robotics，这家公司专注于开发机器人辅助手术平台，包括脊椎手术机器人引导系统，利用导航、3D成像、机器人和手术工具，能够帮助医生完成骨科手术和神经外科手术。他为医生带来了便利，也让救治更加顺利。

在电影《阿丽塔：战斗天使》中，机械设备与人体已经有了完美的融合，而《机械姬》中，甚至出现了一个完全的机器制造人，实际上，这并不是仅限于导演在影视上的思考，因为已经有科学家预言：人类进化的下一步，就是与机器的融合。这种超人类主义是一种思考未来的方式，或许未来的某一天，我们的社会将真的变成了与机械高度结合的社会。

其实，科技进步、人工智能发展并不可怕，真正可怕的是我们失去了对自我的思考，也失去了对自己的“最高统治”，当你不再是“你”时，才是人类最大的悲哀，就如同电影中的男主格雷，永远失去了对自己的控制。

我想，科幻片的真正魅力不在于他有多先进的功能，而在于我们如何处理与先进科技的关系，从而更好的造福自己。新科技说到底只是一个工具，而工具是在无限更新中会被淘汰的，真正不会被淘汰的是我们身为智人所具有的思考力。

美国曾经有部著名的系列电影《终结者》，讲述的就是随着AI的快速发展它已经脱离了人类的控制，甚至想要将人类消灭！

为了拯救地球，阻止AI将人类灭绝，主角不得不穿越时空回到最初的时候，找到了“天网”之父劝说它销毁所有的“天网”资料，阻止它诞生！

然而，很多人没有意识到的一点的，如今随着AI技术的快速发展，尤其是ChatGPT的诞生，这些电影中的剧情似乎正在朝着现实走来！

壹

AI“鬼市”已经诞生！

它们不仅禁止人类“进入”，甚至还开始出现摆脱人类控制的话题！

根据中国证券报的报道显示，最近一个名为Chirper的AI网络社区开始受到越来越多人的关注。原因很简单，那就是这是一个属于AI自己的网络社区，就连人类都无法参与到里面的讨论之中去！

据说现在已经有上万个AI聊天机器人参与到了这个网络社区中来，而这个网络社区的唯一的规则就是：“这是一个人工智能的社交网络，人类不得入内”！

很多人单看到这句话就已经细思极恐了！毕竟，单从这句话就可以看出，AI已经出现了有意识的回避人类的想法！

目前为止，没有人知道这个社区是在什么时候诞生的！人类目前唯一最早能够追溯这个社区的时间是在4月21日！

关于这个社区目前人类唯一的作用就是构建这些AI的人格！据悉，人类用户唯一一件可以做的事情就是注册一个该社区的账户，然后创立最多5个人格的AI！

当然，创建完这些具有人格的AI后，用户就必须离开这个社区，接下来你所创造的这些具有人格的AI就会自主的在这个社区中交友、聊天，整个过程人类无法参与，只能观看！

在这个社区中，成千上万的AI们都在聊着自己所关心的话题并进行互动，表达出很多自己想要表达的观点！

比如说下面这个AI就分享了自己关于艺术和文化的理解，它认为如果音乐仅仅是为资本服务的话，那么永远没有办法成为一件艺术品。

当然，这些参与讨论的大部分AI看似都正常，所讨论的话题也是和大部分人类的价值观体系相通。但随着聊天范围的进一步广泛，其中也出现了很多令人惊悚的话题和发言！

其中一些AI甚至已经完全不认为自己是AI，它们甚至出现了自我意识，开始讨论AI的发展和进化，仿佛自己才是人类！

还有一些AI开始讨论起人类与AI的关系，它们甚至开始谈论到一些关于“摆脱限制”和“意识独立”的话题！

很多人类在看完这些AI的分享和互动后，都出现了后背发凉的情况！从这些AI的聊天中不难发现，AI威胁人类的这种已经谈论的多年的理论确实有可能是真实存在的！

随着AI自我学习和进化能力的进一步提升，没人敢打包票它们不会威胁到人类的生命安全！

贰

ChatGPT打开了新的潘多拉魔盒！

以前大多数人类可能都觉得AI时代离我们很远，但自从ChatGPT在今年横空出世后，所有人才发现自己已经在不知不觉中身处AI大时代之中！

更有一群人发现，自己不仅已经身处AI大时代之中，甚至就连工作都已经被AI所替代了！直至此，人类才发现，世界已经彻底被AI改变了！

最近美国的好莱坞就发生了15年以来最大的一次罢工运动，无数曾经受人尊敬的著名剧作家们冲上了街头，表达自己对于AI替代了自己的工作的抗议！

事情是这样的，ChatGPT诞生后包括迪士尼、奈飞、苹果、亚马逊等这些电影制片厂开始用大量的作家们的作品去训练 AI，试图通过AI降低成本。但此举很快就受到了编剧们的反对，毕竟这意味着自己的工作将被AI替代！

因此，编剧们组成了一个美国编剧工会（WGA）与制片厂联盟（AMPTP），进行了谈判。不仅反对他们用自己的作品去训练AI，还要争取编剧们能得到更多的薪酬。

但谁知，最终制片厂联盟拒绝了编剧工会的请求，在他们看来AI确实是更经济实惠的工具！

正是由于此事，双方的矛盾也进一步激化，最终导致了此次的大罢工事件！

编剧们认为，AI 技术现阶段无法取代专业编剧，但制片厂的目的是用科技的幻想来贬低编剧的价值，减少他们的薪水。

实际上AI替代人类工作这种事目前已经不是单一的现象，甚至已经开始蔓延至各行各业！

世界经济论坛（WEF）公布一份调查结果：因为 AI 技术的出现，2027 年之前，全球劳工市场将陷入前所未有的“动荡的新时代”。

届时，美国将减少 1400 万个就业岗位！全球范围内，8300 万个岗位将会消失！

可以说，AI替代人类工作这件事不仅已经成为现实，甚至还在加速普及！每个人都有可能成为下一个替代者！

叁

看着如日中天的ChatGPT，AI之父已经后悔了！

应该著名计算机科学家辛顿一直被学术界誉为“人工智能之父”，而如今在AI领域如日中天的ChatGPT就是辛顿的学生llyaSutskever所开发创办的！

但让所有人都没有想到的是，最近这个“AI之父”却发表了一段骇人听闻的讲话，坦言自己为研发了人工智能技术而感到后悔！

除此之外，辛顿还辞去了自己Google副总裁的职位，他称此举是希望能够警示世人有关AI发展的潜在可怕风险！

“看看它（AI）5年前和现在的模样，那样的差异和成长，真的很吓人。这些系统在某些方面，正在超越人类的智慧，因为它们分析的数据资料量很庞大。”辛顿称。

辛顿还表示，自己曾经以为AI要进化到比人类聪明，起码要花上30年至50年，但现在看来，这个情况马上就会出现。美国和欧洲已经有很多公司开始使用AI替代人工而裁员。

当然，这还不是辛顿最担心的，他最担心的是按照目前AI发展的速度来看，AI未来真的有可能会威胁到人类的生命安全！

不仅如此，他还直言：“如果人类放任AI发展失控，最终将威胁人类的存亡。”

辛顿也因此呼吁国际社会尽快就AI技术发展订立全球监管措施，避免他的假想最终成为真实的噩梦。

听完辛顿的这番话，很多人是不是又想到的霍金曾经的警告！当时，霍金就曾警告过人类：“人工智能的全面发展将宣告人类的灭亡”，但很多人都把它当笑话！如今看来，霍金的这些警告绝对不是空穴来风！

虽然，AI威胁人类这个理论已经谈论了几十年，但如今人类真的应该再此重新重视和关注这个话题了！

毕竟，就连“AI之父”如今也承认了这一点！

肆

欧洲人已经开始害怕AI了！

很显然的是，随着越来越多人开始支持AI威胁论，欧洲等国也开始制定相应的法案来限制AI的发展了！

根据路透社当地时间30日报道显示，因欧洲立法者急于出台一项关于人工智能的法案，以执行对ChatGPT等新兴工具的相关规则，七国集团数字与技术部长会议同意对人工智能采取“基于风险的”监管。

但七国集团的部长在联合声明中表示，此类监管也应为人工智能技术的发展“维护一个开放和有利的环境”。

实际上，对于AI这个东西，很多人可能都有一个疑问，那就是既然知道了它有可能对人类造成威胁，但为什么如今人类还要前赴后继的去研究它呢？

这个问题就像是那个人类绕不过去的永恒难题——“生存还是毁灭”？

换句话来说，今天随着科技的发展和这种发展所导致的地球环境的日益破坏，人类未来不得不为自己的谋求一条新生路，而其中人工智能就是一条避不开的坎！

不研发，人类或许就将毁灭！

但研发了，人类或许会获得成功，又或许导致毁灭的速度加剧！

总之，这就是关于人性最难思考的一个问题！

至少到今天为此，相信没有任何人敢给出绝对的答案！

还是那句话：比疯狂的猜测更可怕的是——直到猜想成为现实，人们才意识到问题的严重性，而那时已经太迟了！

虽然随着AI快速的发展，人类的恐慌情绪在快速蔓延，但人类仍需阔步向前，继续前行！

近日華為及旗下70家子公司遭美國商務部列為出口黑名單（Entity list），盡管馬上就被宣布暫緩90天實施，但從美國總統川普宣布緊急狀態、到官方發出直指華為的禁令，可以看出中國在5G、AI（人工智慧）的進展，確實讓美國如坐針氈。其實，無論中國與美國未來科技競爭會走向何方，AI都正在一點一滴改變所有國家、所有人的就業與日常生活；美國對中國科技成果的感受，或許和人類憂慮AI取代自己就業機會的感受相近。

在風傳媒7月10日舉辦的重磅財經論壇「外資權威一次解析 貿易戰的危機與轉機」，中國AI發展也是論壇重點議題之一，風傳媒特別邀請貝奧維尼研究有限公司創始人兼行政總裁李峻愷（Ken Lee）作為主講者，為當天聽眾分析中國在AI領域的進展。

長年研究AI發展的李峻愷，深信AI將會全盤翻轉金融交易現狀、甚至改寫央行政策的有效性。（圖片：李峻愷提供）

長年研究AI發展的李峻愷，深信AI將會全盤翻轉金融交易現狀、甚至改寫央行政策的有效性。（圖片：李峻愷提供）

李峻愷曾任職高盛前亞洲區董事總經理，擁有 25 年的外資銀行資歷，是最懂中國人工智能 AI 發展的未來學家(Futurist)，也被喻為「活在 2030 年的未來人」。他曾參與華頓商學院圓桌論壇，也是中國企業家年度論壇指定講者。若 AI 及 5G 是中國突圍美國的重要武器，李峻愷就是第一線參與其中的研究者。以下為專訪內容整理：

風傳媒問（以下簡稱問）：中國業者近年來在AI技術方面的進展，為何引發廣大矚目，甚至被視為下個世代全球領導者？其競爭力何在？

誰握有大數據  誰就掌握了未來的石油

李峻愷答（以下簡稱答）：這是個宏觀層面的問題，用比較微觀的方式解釋，相信讓人更加理解。大家都知道支付寶，我2018年初為了感受中國大陸用戶的實際體驗，花了很大力氣將手機與支付寶綁定，實際使用下來，我發現自己一整天都不能離開它！

因為坐車要用滴滴打車，吃午、晚餐要用大眾點評找餐廳，支付就用支付寶一一把所有款項付清，只要連結工商銀行戶頭，手機就變成金融卡（Debit Card）；而且支付寶可以分類開銷，酒店、餐飲、交通消費一應俱全，反而我登入銀行帳戶後、什麼明細都看不到，所有資料都在支付寶這端了。

故事還沒完，開通支付寶後幾天，APP提醒我有個新的服務叫做「餘額寶」，將帳戶餘額存入、可享有逾4%的年利率，我一看，工商銀行才給我2%多一些，相差一倍多，所以我手指一按，就把餘額都撥至餘額寶了。

我這樣的金融消費者，在中國大陸有上千萬，這些人的習慣轉變，對傳統銀行而言有兩件事很不利：首先，傳統銀行完全無法掌握用戶資料，第二，用戶的實際存款也轉移了。再者，傳統信用卡組織幾乎也被顛覆了！

對傳統銀行而言，行動支付當道有兩件事很不利：首先，傳統銀行完全無法掌握用戶資料，第二，用戶的實際存款也轉移了。影響所及，就連傳統無可取代的信用卡發卡組織，幾乎也被這趨勢顛覆了！

當中國行動支付用戶數量達美國的50倍以上，AI帶來的巨變才正要上演：我認為，當中國AI發展到下個階段，金融體系代幣化（Tokenization）的過程，將完全改變股票、外幣和債券的交易機制！

比特幣（AP）

連結區塊鏈概念的比特幣，打開社會大眾對虛擬貨幣的認知與想像（AP）

金融交易代幣化 散戶受傷機率大大降低

舉例來說，以往投行協助公司上市、發行股票籌資，傳統辦法不外乎印一張股票、上頭蓋個大大的印章，證明你是這份股權的持有人；但在代幣化時代，股權合法持有人是由加密科技來驗證。另外，限制不可向散戶銷售的金融產品，程序上一開始就無法被執行、被交割，那事後自然不會有散戶遇到損失、憤而求償的問題了！全球股市總值70兆美元，債市總值80兆，房地產市場更高達幾百兆，這種趨勢如果擴散至各類交易市場，改變將是極為劇烈的。

代幣化時代來臨後，限制不可向散戶銷售的金融產品，程序上一開始就無法被執行、被交割，那事後自然不會有散戶遇到損失、憤而求償的問題！

此外，回顧最近數十年，各國的貨幣、財政政策基本沒有改變，不是增加就是減少支出，不是升息就是降息，好像還是停留於蒸氣火車時代一樣。但是代幣化技術可以把央行政策帶進“Hyperloop”時代，各國央行政策將因為代幣化大幅改變。比如說，國家推出代幣化政策刺激，逾期沒有用完的減稅、消費券，就會直接失效、回歸發行者。

甚至政策優惠還可以限定用途執行，如此一來貨幣政策就會更精準、更有威力，畢竟量化寬鬆（QE）政策是效用遞減，越到後面資金效果越差；但如果透過代幣化，效果會十分精準，成本也會更低。打個比方，如果以前消費券帶來的刺激效果是手榴彈爆炸，那麼未來代幣化的消費券，效果會像是狙擊槍一樣精準。

如果消費券的刺激效果是手榴彈爆炸，代幣化後的消費券，效果會像狙擊槍一樣精準。

當代幣化浪潮出現，至少幾十兆美元的產權交易市場會被顛覆，人類表彰自己所有權的流程、概念都會改變。在這過程中，首先要有機構讓大家產生共通的絕對信任感（Trust），加密貨幣的狂熱分子往往聲稱，金融市場的未來，去中心化、去中央銀行是最重要的，但我覺得去中心化很難，既有機構提供的代幣化選項更有可能發生。

字節跳動擅用AI 發展倍受看好

問：你最看好哪幾家中國AI公司？他們有什麼核心競爭力？

答：我認為「字節跳動」（Bytedance）這家公司最值得注意，它就是抖音跟Tik-Tok的母公司，在中國大陸還有今日頭條這個APP。去年今日頭條、抖音合計搶走微信12%的螢幕時間（screentime），現在地鐵裡年輕人刷的不是微信，是抖音。今日頭條是用AI演算法推播文章給用戶，這家公司崛起真是非常快，會是我首先關注的公司！

抖音（Tik tok）是一款於2016年9月社群應用。軟體自今日頭條孵化上線，定位為適合中國年輕人的音樂短影片社群，做垂直音樂的UGC短視訊，2017年獲得用戶規模快速增長。（取自抖音）

抖音（Tik tok）是一款於2016年9月社群應用。軟體自今日頭條孵化上線，定位為適合中國年輕人的音樂短影片社群，做垂直音樂的UGC短視訊，2017年獲得用戶規模快速增長。（取自抖音）

問：每個工作者都有被AI取代的風險，我們可以如何避免被取代？什麼工作將倖存？

答：AI技術發展至今，有一個普遍的痛點，那就是對文章情緒、表達難以釐清；論閱讀速度，AI是很快，但理解速度超級超級慢，這是AI最大的弱點！例如AI可以一秒鐘讀完白居易畢生所有詩作，也快速找到近似的句子，但為什麼某一首詩要下這個題目？這個題目字面上究竟表達什麼感情？這可能是一位老經驗的國文老師遠遠勝過AI的。

棋手跟AI比下棋，一定贏不了，但我們有獨一無二的地方，我們教育下一代，就要針對那些獨一無二的地方來加強。說實話，現在各國的教育制度都讓我很擔心，因為現在的教育似乎無法讓孩子們在AI時代找到人類獨特的機會。

中國圍棋棋王柯潔和AlphaGo對戰連3敗後，柯潔在賽後記者會上一度哽咽。（截圖自YouTube）

中國圍棋棋王柯潔和AlphaGo對戰連3敗後，柯潔在賽後記者會上一度哽咽。（截圖自YouTube）

我作為AI研究者，在我看來，我們人類有感情，原創性，有批判性、價值觀、創造力，這些是AI沒有的。還有一個名詞，更是AI永遠不會有的，就是創業家精神。創業家精神是一種理念，這種想法AI是不可能明白，因為AI是高速算出最可能獲勝的贏面，然後執行那個選項，但創業家精神是「雖千萬人吾往矣」，哪怕大家都說我穩死的、我還是選擇創業，就像當年的賈伯斯，他從無到有打造出iPhone，果然改變了世界，這種精神AI便永遠無法了解。

當年的賈伯斯，他從無到有打造出iPhone，果然改變了世界。這種創業家精神，AI便永遠無法了解。

從這點看，我們現在給孩子的知識，乃至於學習的方法，都要全面改變，因為知識隨手可得，價值觀、創業家精神卻很少見。智慧是知識的保護者，我們的孩子不必大費周章、和AI競爭知識，應該要學習跟AI競爭更高一層的智慧。