In the next 30 years, artificial intelligence may bring too many social phenomena that human beings do not understand at present. Failure to understand may bring troubles or even disasters. The first is to replace human work and even make some people lose the meaning of ” being born”.
When artificial intelligence starts to move from laboratory to wider application, it will no longer only have technical impact, but will more and more obviously affect the operation of human economy and society. Artificial intelligence not only frees human beings more, but also brings negative effects such as unemployment and widening income gap.
Besides, it came so fast. After everything is networked, artificial intelligence can make all kinds of more accurate decisions, judgments, predictions, and classifications than people as long as it uses numbers to calculate and push.
I also made a systematic analysis of this topic in my new book AI. Future: ” Traders on Wall Street, this once glamorous profession soon disappeared; The future security will also disappear partly, because the surveillance of cameras and the inspection of some robots no longer require security. Drivers may also disappear, including some very high-end white-collar workers, such as radiologists, who are not as good as robots. ”
Why is this result? The analysis of artificial intelligence replacing jobs cannot be carried out in a single dimension of traditional ” low skill” versus ” high skill”. Artificial intelligence will produce both winners and losers, depending on the specific work content. Although artificial intelligence can outshine human beings in a few jobs based on data optimization, it cannot naturally interact with human beings, nor is its body movements as dexterous as human beings, nor can it do creative cross-domain thinking or other jobs requiring complex strategies ( because the elements and results of these jobs cannot be easily quantified ). Some jobs that seem difficult to human beings may be very simple in artificial intelligence. Some jobs that seem simple to humans may be the death throes of artificial intelligence. We can use the following two figures to illustrate:
For manual labor, the left side of the x axis is ” low skill, structured” and the right side is ” high skill, unstructured”. Under the Y axis is ” weak social contact” and above is ” strong social contact”. The Y – axis of the mental work graph is the same as that of physical work ( weak social interaction to strong social interaction ), but the X – axis is different: the left side is ” optimized” and the right side is ” creative or decision – making”. If mental work focuses on maximizing quantifiable variables in the data ( e.g. setting optimal insurance rates or maximizing tax rebates ), it is classified as an ” optimized” occupation.
These axes divide the two maps into four quadrants: the third quadrant is the ” danger zone”, the first quadrant is the ” safety zone”, the second quadrant is the ” junction zone” and the fourth quadrant is the ” slow change zone”. Work that mainly falls in ” dangerous areas” ( such as truck drivers ) will face high risks of being replaced in the next few years. It is unlikely that the work of ” safe areas” ( such as psychotherapists, physiotherapists, etc. ) will be automated in the foreseeable future. The boundaries between the ” integration zone” and the ” slow change zone” quadrants are not very clear: although they will not be completely replaced at present, the reorganization of work tasks or the steady progress of technology may lead to large-scale layoffs for these jobs.
In the ” junction area” in the upper left corner, most of the calculation and physical work can already be done by machines, but the key social interaction part makes them difficult to fully automate. Therefore, the most likely result is that behind-the-scenes optimization is done by machines, but human employees are still required to be the social interface of customers, and human beings and machines form a symbiotic relationship. Such jobs may include waiters, financial advisers and even general practitioners. The speed and proportion of these jobs disappear depends on the flexibility of the company in transforming the work content of its employees and the openness of customers to interact with computers.
Jobs that fall into the ” slow change zone” ( such as plumbers, construction workers, art designers, etc. ) do not rely on human social skills, but rely on flexible and ingenious craftsmanship, creativity, or the ability to adapt to unstructured environments. These are still the short boards of artificial intelligence. As developing technologies will slowly upgrade these short boards in the next few years, the speed of work disappearing in this quadrant depends more on the actual expansion of artificial intelligence capabilities.
In my opinion, warnings, pessimism and panic are groundless worries that ” do not know the true face of the mountain.” Tearing off the label, artificial intelligence is neither ” human” nor ” intelligent”. It can only become a human tool and cannot replace all human work. For workers, the pressure from robots and artificial intelligence algorithms is huge, but the prospect is not completely bleak. Artificial intelligence will replace many jobs currently performed by human employees. In some industries, this trend has begun to show, but some jobs cannot be replaced in the foreseeable future.
The following are the 10 jobs I predicted to be most vulnerable to AI attacks and ” in name only”.
1. Telemarketing Staff / Telemarketing
Telemarketing will be one of the fastest jobs to be replaced by AI. You have probably received automatic voice calls, which will become more and more natural in the future. In a single-domain dialogue dominated by AI, AI is also easier to achieve real results. In addition, AI will find ways to attract customers through customer information, purchase history and facial expression recognition. For example, a gentle female voice or a persuasive male voice can be used to make up-selling to impulse buyers, and goods with appropriate prices and categories can be used to target customers. Compared with manual telemarketers, AI has almost zero cost, no complaints, high performance and high consistency with business logic, so telemarketing jobs have no future. If you are engaged in telemarketing, you can consider changing to face-to-face sales or other work that requires dealing with people.
2. Customer Support
Customer interaction will increase with the application of AI. However, due to the repetitiveness of this type of work ( there is usually a textbook response method as a reference ), customer support will be largely replaced by AI. This process will be carried out in several stages. The first to be replaced will be chat robots and mail customer service, followed by voice services involving a large number of incoming calls and relatively simple products / services. In the beginning, AI will work together with human beings, and AI will provide suggested answers, themes and fixed replies. Humans will act as backup personnel to handle calls that AI cannot handle ( e.g. callers are angry ). This will shorten the waiting time of customers, improve the problem solving rate ( because the premise of using AI is to confirm that it can solve the problem ) and greatly reduce the cost. This process will accumulate a large amount of data for AI and eventually make AI perform better than human beings. If you are engaged in customer service, you can change from text support to voice support, from light support to deep support, from telephone / Internet service to face-to-face service. At the same time, you should also learn empathy, communication and persuasion skills.
3. Warehouse workers
Amazon warehouse has adopted robots developed by Kiva system. They will move shelves to the front of human workers in fixed positions, who will pick up goods and put them into boxes. However, with the development of computer vision and robot control technology, the work of fixed-position workers will experience the fate of increasing intensity and then being replaced. In addition, AI will soon be able to move boxes, load cars and other warehouse work. Compared with factories, warehouse automation requires less precision and is therefore easier to implement.
4. Cashier and Operator
Citibank recently announced that it will reduce the number of operating personnel from 20,000 to 10,000. These are ” nameless” middlemen who deal with data and information. Their responsibilities include document filing, processing, purchasing, inventory management, error investigation, sales estimation, and reporting the results of investigations to management. With the electronization of business process, business intelligence system can automate the whole process and AI can even make decisions directly. This phenomenon will occur not only in the banking sector, but also in every large company dealing with massive data. In the AI era, no one wants to be a data processor with a thousand people.
5. Telephone operator
The telephone operator is the job that requires the least interpersonal skills. Now speech recognition is becoming more and more accurate ( Microsoft’s speech recognition has surpassed the human level ), and speech synthesis oriented by situational dialogues is becoming more and more natural ( Google’s latest speech synthesis is almost indistinguishable from human voice ). In addition, telephone work is also challenged as more people rely on information, so it is only a matter of time before it is completely eliminated.
6. Cashier / Cashier
Cashiers and revenue collectors are being replaced by ATM machines and self-checkout machines. The increasingly fierce competition has forced retailers, banks and fast food companies to drastically streamline their labor processes. Amazon Go Unmanned Store has heralded a future in which stores are completely unmanned. However, due to the high price of unmanned stores, the lack of popularity of mobile payments and the privacy problems of cameras and facial recognition, unmanned stores will not be quickly mass – laid. But if you are a revenue collector, don’t feel relieved. Self – checkout machines based on radio frequency identification ( RFID ) and computer vision are coming fast, along with smart vending machines and small convenience stores. It is time to change jobs as sales assistants, preferably those that evaluate and reward affinity and persuasion.
7, fast food clerk
Food preparation has the characteristics of repeatability and fixed location, so it will inevitably be replaced by AI. Existing chain restaurants have begun to promote automated ordering processes, and face recognition and voice recognition technologies are likely to be used soon. The next step is naturally to automate food preparation and cooking. In addition, there will be a new affordable chain restaurant with full automation of cooking and serving in the future ( for example, a robot Chinese restaurant with half the price of McDonald’s ). These ” robot restaurants” will take away the business of the traditional fast food industry, resulting in a decrease in the number of fast food employees.
Instead of thinking of the dishwasher as a robot, think of it as a super-large dishwasher that can remove dishes ( and of course food, bones, napkins and other tableware ) directly from the table and then polish the dishes and silverware. Dishcraft, a California – based start – up, is already selling this super-large dishwasher. These dishwashers are indeed expensive, but for large restaurants, they are still acceptable compared with the saved labor costs. Over time, large-scale production will reduce the cost of dishwashers. If you are currently engaged in repetitive work such as washing dishes, it is time to receive training and change to a less repetitive job.
9. Quality inspector of production line
Production line work will be phased out gradually, with high repeatability and fixed working environment. The whole elimination process may take as long as 20 years, because it is still difficult for AI to control robots. However, there are also some easy-to-use production line work for AI, such as checking the damage and defects of goods ( such as checking the iPhone case to ensure the beauty of the product, or checking the circuit board to ensure the function of the product ). This kind of work takes advantage of the rapid development of computer vision and requires little or no control. For human inspectors, this kind of work is troublesome and tiring, especially hurting eyes. Therefore, it is time to change to a job that is less harmful to health, requires high sensitivity, and needs to be done in a new environment with non-fixed structures.
Couriers and couriers are being replaced by courier robots, small cars, large trucks and unmanned aerial vehicles. The first will be indoor distribution services ( hotel rooms and apartments ) in a structured environment, then will extend to non-public roads, and finally penetrate the entire express delivery industry. In the short term, e-commerce will continue to grow and the demand for express delivery will also increase. However, express delivery is by no means a good choice, involving very little human expertise and interpersonal interaction.
Worries about the replacement of human jobs by intelligent machines are not unfounded. From truck loaders to legal researchers, no matter white-collar or blue – collar, as long as it is repetitive work, robots and software based on artificial intelligence will replace it. In the era of intelligent machines, what is the meaning of human beings?