Four years, for this wave of artificial intelligence, has been enough to trigger a vicissitudes of life.
I remember in 2016, in the auditorium of a university in Beijing, I heard Wu Jun, a well-known science writer and Silicon Valley investor recommending his new book "Smart Times" at the time. He made bold predictions on stage: only 2% of people can cross this intelligence In the revolution of the times, the remaining 98% of people may be those who have been replaced or eliminated by artificial intelligence.
At the time, I was deeply skeptical of this conclusion. After the sharing, I squeezed onto the stage and asked Dr. Wu Jun this question: If 98% of people are destined to be eliminated, how will these people live in the future?
In terms of time, Dr. Wu Jun has not been able to give too much explanation. He is more optimistic and said that in the future, it may depend on strong government control, and the rich will pay more taxes to increase the welfare of other classes. Regardless of whether this vision of an ideal society will emerge, the rapid advancement of artificial intelligence technology in the following years has indeed exceeded our expectations, and the trend of "intelligence" instead of "artificial" has become more apparent.
Last week, I participated in this year's Beijing International Information and Communication Exhibition. I witnessed the innovative revolution of 5G network and artificial intelligence technology to the traditional industrial working methods, such as unmanned workshops, remote operation of port freight dispatch, and autonomous driving delivery. More and more scenarios are appearing in the replacement of artificial labor by AI robots, or because of the application of artificial intelligence technology, the on-site operation and maintenance personnel are greatly reduced.
We see that this trend has caused two results. One is that many people’s jobs have been replaced by AI, and these people have to make a living; the other is that AI has replaced many repetitive and dangerous jobs. Objectively increase the labor welfare of workers.
Obviously, no matter what kind of consideration, the development trend of artificial intelligence technology will only get faster and faster, and we have to face the threat of artificial intelligence technology to replace our jobs, and we must also think about large-scale artificial intelligence technology. After taking up the post, it has both positive and negative effects and influence on human laborers.
Under the epidemic, a large wave of AI robots are on duty
The emergence of the new crown epidemic this year has led to the acceleration of two trends, one on the left and the other on the right: one is the ebb of globalization, from open exchanges to mutual closure, fighting each other, and a substantial reduction in the flow of transnational personnel; the other is the acceleration of the digital economy. The physical isolation caused by the epidemic has triggered the digital upgrade of traditional industries and the growth of online office models. Of course, this trend has already appeared before the epidemic.
Simply put, it is to minimize human contact in the physical world, and increase human-computer interaction and interpersonal interaction as much as possible in the digital world. Whether it is personal intuitive experience or trends seen in media reports, we can deeply feel that among digital technologies, AI technology is innovating existing working methods and replacing and dominating existing workers.
The emergence of the epidemic provides an excellent opportunity for this kind of AI robot to get on duty. For example, some farms in the UK and Australia are welcoming large-scale ripening of fruits and vegetables during this season. Originally, these farms would hire seasonal pickers at this time, usually workers from Eastern Europe and backpackers from all over the world. Completed the task of picking, but affected by this year's epidemic, these people have difficulty entering the country, which once caused a very serious labor shortage.
Some farms in the United Kingdom offer a salary of 500 pounds a week, and it is difficult to motivate local young people to engage in these hard and boring jobs. Some farms in Australia even offer super high salaries of 3,800 US dollars per week for recruitment, which is higher. The average income level of Australian workers is three times, and there are also no young people willing to work.
On the one hand, there are agricultural products rotten in the ground, on the other hand, the prices of fruits and vegetables in supermarkets have risen. Farmers in some places have begun to actively seek the help of AI robots to replace manual picking by picking robots. According to reports, the UK has launched Robocrop, a robot that can pick raspberries. It has four claws and is guided by sensors and a 3D camera to perform picking at the same time. Robocrop can pick for 20 hours continuously, and can pick up to 25,000 raspberries per day, while the eight-hour shift workers can pick a total of 15,000 raspberries per day.
At present, the cost of picking robots is high, and the picking speed is slightly lower than that of manual labor. However, if the method of renting and continuous work is adopted, the picking cost will be lower than that of hired labor.
If this picking robot model is successfully promoted, it will inevitably bring about the following two consequences: When farmers find that the automated picking of AI robots will bring more significant cost reduction and efficiency increase, then they will be more involved with these Robot R&D institutions cooperate with companies to promote the standardized and automated process of planting and picking. At the same time, these farms no longer need the labor of foreign workers in the future, and consequently these people will lose a large part of their income.
Another is that due to the impact of the epidemic, the decline in business profit expectations of companies has led many companies to take the initiative to lay off employees or cut human expenditures. Instead, artificial intelligence technology solutions are used to replace these human resource expenditures. This phenomenon can be seen in a large number of cases from the high-end banking and financial industry, to the automobile and other manufacturing industries, and even the Internet industry.
At the beginning of the year, Microsoft replaced a large number of human editors with AI editors, and it also surprised a lot of news media people. At the beginning of the year, YouTube also abolished the global content manual review team on a large scale, and replaced all of them with AI for content review. This move saved YouTube a lot of manpower and made YouTube the first time in history. Throughout the quarter, no human auditors participated in the preliminary review of content.
The emergence of this series of "intelligent" versus "manual" alternative scenarios does not occur under the epidemic, but only because the arrival of the epidemic has become more urgent and the process of accelerating the replacement has become faster. Now, we have to think about the possibility of our work being replaced by AI, or the threat of being dominated by cooperating with AI in the future.
Or take the initiative? Or passively eliminated?
Why do we need artificial intelligence? At the national level, the development of artificial intelligence is a country’s core technological competitiveness, and its importance is unquestionable; for governments and corporate organizations, AI technology is an effective technical means that can be widely used in social management and enterprise production management. It is also a representative of a new type of productivity that can create huge social and economic wealth. For us individuals, AI technology can provide us with a safer and more convenient living environment, and provide a variety of efficient and abundant intelligent solutions for life and work.
While we have tacitly enjoyed the benefits that AI technology brings to us, we must also see the alternative threats that AI technology brings to our work value.
The first thing that bears the brunt is that AI technology has been added to the work system that collaborates with human workers, and has begun to dominate or control workers. A typical case is the recent article "Takeaway Riders, Stuck in the System", which revealed a safety dilemma of "races against time and life" caused by algorithm lock-in of takeaway riders.
Due to the layoffs of many companies and the suspension of the service industry due to the epidemic, a large number of employees have poured into the takeaway industry with the lowest employment threshold and low skill requirements. The survival consideration of obtaining a basic income has become their first consideration. In this way, the platform hopes to maximize performance, consumers want the least waiting time, and riders want to send more orders. The algorithm is constantly being trained to be efficient as the optimal goal. Naturally, the rider’s meal delivery time will be "optimized" to the extreme. This leads to such a dilemma that all parties can shirk their responsibilities, but are ultimately difficult to resolve.
Repetitive, structured, and time-rigid tasks like takeaway riders, blue-collar workers such as sorting workers in e-commerce warehouses, express drivers in shared travel, and front-line workers in assembly lines, are the most likely to be optimized and transformed by AI algorithms. In the field, on the one hand, they have to accept real-time monitoring by AI systems, cameras and user evaluation systems. On the other hand, companies are still studying corresponding AI robots to completely replace their jobs.
If you think that only these blue-collar professions are under the "supervision" of AI robots, then "Too Young, Too Naive". According to a report in the Wall Street Journal, during the epidemic, as a large number of companies in the United States started working at home, more and more companies began to adopt monitoring management software to automatically record and analyze employee operation records on computers. , And can also take screenshots as evidence. What could have been an intelligent tool that could optimize the work efficiency of employees went to the opposite side of itself, becoming a "restless" supervisor and "informer".
The more direct threat lies in the direct replacement of a large number of job workers by AI algorithms and their robotic equipment.
For example, autonomous driving is crossing the critical point of large-scale commercial use. Recently, Waymo has received attention again, because this time Waymo has really removed safety officers from some autonomous taxis, and has achieved true driverless travel. Almost at the same time, Baidu Apollo also opened up self-driving taxi taxi services in Beijing.
We always feel that autonomous driving is still a long way from entering daily life, and there is still a long way to go to replace tens of millions of drivers. In fact, it does take a long time to play with the technology iterations, hardware popularization, infrastructure and legal system in the early stage of autonomous driving, but once these conditions are mature, its popularization speed will hit like a tsunami. At that time, for most private car owners, this means the complete liberation of time on the road, and for truck drivers and taxi drivers, this is tantamount to the disappearance of job opportunities.
Similarly, for those mental tasks that are highly structured, procedural, and weak in social connections, such as bank counter staff, stock traders, telephone customer service, and even professional positions such as lawyers, doctors, and financial accounting who deal with basic documents, All have encountered the threat of being replaced under the upgrade of specialized AI systems.
We can use this standard to review our current working mode, and reflect on which links are currently cooperating with artificial intelligence, which links are likely to be optimized by AI, and which links have been replaced by AI systems.
For example, in the past two years I saw a cashier in the supermarket trying to introduce self-settlement and facial recognition payment technology to customers. Now, many people can easily check out and pay independently, and the cashier only needs to open a few windows to check out the elderly customers. Obviously, this process has reduced the number of cashier staff.
For example, we media people, in the process of data collection and writing, there are already a large number of search algorithms and input method algorithms to help us improve efficiency, but we still need to do it ourselves from the concept of topic selection and layout. However, GPT-3, which is in front of us, can already start with a topic or sentence, and can edit a complete long article in a decent way. If we compare the output, we will be easily tortured into scum, but if we are more readable, we may still have a little chance of winning. After all, which human is going to read the inhumane articles written by these inhumane machines.
In the current situation where AI technology is aggressively invading our work, we really can only have two attitudes, either choose to actively accept and cooperate with AI, or wait for passive elimination and defeat in the face of AI's outstanding performance.
However, how should we view this "human abdication, AI on the throne" process?
"As soon as humans complain, AI laughs"
Milan Kundera said, "When humans think, God laughs." Borrowing this sentence, we can extend the sentence "As soon as humans complain, AI laughs." How should this be understood?
If you observe carefully, you will find that the introduction of AI technology in human production and life is related to people's various dissatisfaction and complaints in all aspects of work and life. For example, if the driver complains about getting up early and driving in the dark for a long time, then autonomous driving will appear; the security guard complains that staying up late is too hard, and the intelligent monitoring will appear; the customer service complains that the call is boring and the other party's attitude is rude and unreasonable, so intelligent customer service appears; power workers complain about the wild Mountain patrol is too hard, miners complain that it’s too dangerous to go down, tower crane workers complain that the operation is too lonely, then the corresponding automation equipment and remote control will appear; administrative clerks complain that the work is dull and dull, then intelligent applications and automated process office will appear Up...
We should first look at the replacement value of "AI versus labor" from a positive perspective. Obviously, the emergence of AI has not only greatly improved labor productivity, but also can effectively reduce labor intensity of laborers, improve their labor environment, and solve a large number of repetitive, dangerous and harmful work problems that are harmful to health.
It can be said that since human beings entered the agricultural society in the era of gathering and hunting, they have been engulfed by endless heavy manual labor, but in fact, human nature is still pursuing comfort and enjoyment. With the modern industrial revolution entering large-scale mechanization and automation processes, laborers Have a lot of leisure and rest time again. The popularization of artificial intelligence technology will completely push this process to the extreme, that is, a large number of laborers will be truly liberated.
The by-product of this process is the disappearance of jobs that originally required a large amount of labor, and more and more people have no substantive jobs. The existing salary distribution system of “no laborers can’t eat” can’t cope with this problem at all, and my concern at the beginning of “how do I survive after artificial intelligence replaces my job” is precisely this value system. The next question.
We might as well predict boldly that once relying on technological progress and the efficiency improvements brought about by AI, the total social wealth generated by the digital economy will be much higher than the current level. Therefore, although a large number of people no longer engage in their original jobs, society Corresponding wealth will also be transferred to these people, but obviously most people can only get the benefits needed to maintain a basic life. As the aging society intensifies, this process will become more obvious. On the one hand, the society’s young labor force is insufficient, and large-scale AI robots must be used for production. For example, facing the over-ageing of the agricultural population, Japan can only vigorously promote the automation and intelligence of agricultural machinery to realize unmanned agriculture. On the one hand, the increase in social wealth ensures that these "unemployed" still enjoy a relatively rich standard of living.
For those workers who can establish collaboration with AI systems, this process will take another look. That is to get involved in a fast-paced and efficient collaboration network, and accomplish more complex tasks with the blessing of AI. These people can naturally get a high amount of wealth and a sense of work accomplishment, but at the same time, these people will become more and more busy in this collaborative network, truly becoming the 2% of the population described in "Smart Times".
Going back to the previous story, the takeaway rider caught in the algorithm system dilemma, because there are also various "complaints", the future is also likely to be replaced by another system, such as unmanned delivery vehicles and drones, or even indoors in the future. Replaced by fast food robots with smart printing. However, in the process of the replacement of the old and new systems and professional values, the non-standard and flexible profession of takeaway riders is still a choice for many other eliminated workers to "temporarily transition".
However, YouTube’s AI system, which replaces human content reviewers, is too efficient and flags videos that violate the rules too broadly. As a result, YouTube has to hire experienced human reviewers again. AI still has a long way to go in distinguishing between what is pornographic, violent, and horrible content, and what is art, jokes, and documentary reports. However, for human auditors, this work does have occupational risks that harm physical and mental health. Obviously, such work is more suitable for AI in the future, and human auditors must act as the final gatekeeper based on user complaints.
In the future, we will go hand in hand with AI in all aspects and be able to collaborate with AI algorithms. Although it is hard work, it is valuable. This is a lucky thing. If you can’t collaborate with AI, but you can enjoy enough leisure and entertainment, this is also a good thing. At least we are still constantly releasing our living data as human beings to the AI system to form a larger-scale collective wisdom of mankind . This may be our most important value for the future.
No comments:
Post a Comment