Compared with the bottom workers, white-collar workers are more likely to be replaced by AI.

  When I wrote “Beijing Folding”, I predicted the social impact of robots replacing human labor. At that time, I thought that the biggest impact was the underlying labor force, but according to the current technology trends, the work of primary and intermediate white-collar workers is the easiest to replace.
  Why do you say this way? Because the bottom labor force is easy to be replaced by factory workers, but the service industry practitioners are difficult to replace, because the flexibility of the robot is not as good as human, the non-standard working environment will make the robot feel at a loss.
  But relatively speaking, many white-collar jobs are very suitable for AI because of the simple working environment, duplication of work content, and basically the work of dealing with data and documents. It can be said that as long as it is standardized and repetitive work in the future, most of them can be handed over to AI.
  In the future, what capabilities do white-collar workers need to be replaced by robots? I classify these abilities into the following three types: The
  first ability is generated around the development of AI.
  AI is not a savage beast. On the contrary, they can better make our work more effective. Humans need to understand AI and use AI to better serve themselves.
  With the smart world, basic thinking skills are still important. In the AI ​​era, knowledge and technology are updated very quickly. People need to constantly learn from themselves and let their own update speed match the times. The self-learning ability, the most needed is a good self-reading ability, abstract thinking ability, self-reflection ability.
  In the future, there will be a series of derivative professions around AI. Ordinary people, even if they don’t understand the technical principles behind AI, can fully utilize AI tools to improve their life and work and make society better if they can fully understand its application scenarios.
  The second ability is generated around the field of interpersonal communication.
  For a long time to come, communication between people is still irreplaceable. On the one hand, even if AI develops further, they still have a big gap between understanding the human world and human mind, and thus it is impossible to completely replace interpersonal communication.
  When AI takes over a large number of basic single-type jobs, communication between people will be a more demanding field, and most of the remaining positions and needs may be concentrated in areas where people need a lot of communication and collaboration. Therefore, in order to keep up with the development of the intelligent age, the ability to communicate with people will become more and more important.
  It is conceivable that in the future, it is impossible to do the same work as the past, and a job can be done forever. Standardization work is easy to be automated by machines, not standardization work. It generally means a lot of uncertainty, and it needs to be constantly running, teamwork, communication, modification, adaptability, and mutual compromise. For example, a film crew, some conventional camera production may be automated, some basic scripts and service work can be handed over to the AI ​​each time, but each program still requires a lot of on-site temporary adjustments, communication with the participating guests, programs Collaborative communication, etc., requires collaboration between people.
  In the future, there will be more needs based on human communication in terms of emotional care and escort, and social entertainment.
  The third ability is the key capability of future demand: doing things that are difficult for AI to do, and guiding the AI.
  What is the ability that AI can’t do? The core is two: worldview and creativity.
  The world view is an upgrade of common sense and a panoramic view of the world. At present, AI understands that professional issues are already very good, but comprehensive problems still make them very troublesome. They can’t answer the situational problems in daily life. The problems of everyday life always involve comprehensive common sense in multiple fields of knowledge.
  Human beings have a natural instinct for this. We can construct a model of the whole world and put people on the common sense stage composed of a large amount of background knowledge to understand their behavior.
  The upgrade of common sense gives us insight and a worldview. The richer and more common sense in all aspects, the clearer the relationship between each other, the more you can see the problems of each part at a glance, find systematic solutions, understand the overall situation, and judge the trend. This systematic trend understanding is different from the extrapolation based on past trend experience. It is an understanding of the interrelationship of multi-domain knowledge, and judges the overall trend according to the trend of each part relationship. If you only learn the expertise in a module, it is impossible to grasp the overall situation.
  This ability is more than just a simple input of data. IBM’s artificial intelligence “Watson” entered Wikipedia’s multidisciplinary knowledge a few years ago and defeated human players in the knowledge contest, but it could not use this knowledge to provide valuable innovation for human beings.
  And our worldview can enable humans to have the ability to innovate across disciplines – to make proteomics from the combination of physics and biology; to bring the theory of music into architectural design; to connect political and economic knowledge with life scenarios, and ultimately Presenting the product in the form of pop art.
  Building a panoramic stage of knowledge, combining multidisciplinary knowledge and creating more meaningful things is a step that is difficult for AI to overcome.
  On the other hand, creativity is the ability to generate meaningful new things. It is a combination of multiple abilities that require understanding of old things and imagining new things. Learning and following the existing data is what AI can do, but for the imagination of non-existent things, AI is far less human.
  The new thing that makes sense is because AI currently has a kind of “pseudo-creative”, that is, random production or statistical imitation. With a single program, you can randomly generate 1 million paintings, or count the sentences and bridges in the best-selling novels for imitation and assembly. But this is not a meaningful creation, they do not understand what they have created.
  True creativity should be a deep insight into the problem, a new and unique solution, or a great expansion of the imagination, turning the whimsy into a new, achievable work, or a humanity. Complex comprehension, transforming the unexpressible feelings of the human heart into expressive and touching art.
  Creativity is still the unique ability of human beings. It requires too many human qualities as the basis: aesthetic ability, unique association ability, keen subjective feelings, adventurous spirit, curiosity and self-determination, divergent thinking and convergent thinking. Finally, There is also a strong love for things.