The impact of technological progress on employment has been a social focus and research focus since the Industrial Revolution. The movements and ideologies with the nature of Lutheranism have repeatedly appeared in various forms. However, whether it is from the root cause and the result, or from the essence to the surface, the impact of AI employment this time is really different from before. Everything is predicted before it is established. In the face of potential AI employment shocks, it is necessary to clarify some understanding and establish several policy principles. The impact of AI employment is different from the recurring "technical unemployment" in history. Firstly, this time it is no longer the specter of "technical unemployment" that has repeatedly appeared in history, but a terminator that can replace almost all professions. From entrepreneur Musk, who understands science, to economist Summers, who cares about the development of AI, they all believe that the replacement of jobs by AI will be comprehensive. Once artificial general intelligence (AGI) emerges soon, simple, complex, physical, and intellectual jobs will be inevitable. Secondly, the rapid progress of AI technology is increasingly giving people a feeling of traveling thousands of miles in a day and three autumns in a day. For example, it took 180 years from "Türkiye chess playing robot" (a trick in 1770, which can be regarded as the starting point of this idea) to Turing's paper published in 1950; In 1997, Deep Blue defeated Kasparov and went through another 47 years; It was also about 20 years since the international chess robot named "Alpha Dog" defeated Li Shishi in 2016 and Ke Jie in 2017. And it was only a year since the launch of ChatGPT and the release of Sora. We don't need to use any complex models to predict, just take a look at this speed and acceleration, and we can come up with reasonable expectations for the emergence of general artificial intelligence. Finally, the "development paradox" of large model AI determines that large-scale job losses are almost inevitable. It is recognized among factions, countries, and enterprises that whether they can occupy the commanding heights of AI technology and industry is crucial to their survival. This has led to a competition similar to the space race, arms race, and nuclear weapons race during the Cold War, centered around the development of AI. Moreover, large model AI is highly energy consuming and costly. The inevitable direction and way to explore the use of models, expand user groups, and improve returns is to improve labor productivity, thereby reducing the use of labor and human capital. The two fundamental ways for humans to cope with job substitution have not undergone fundamental changes so far. However, as long as human labor is not completely replaced or determined by artificial intelligence, or until "human-machine integration" is widely realized, there are still some things that will not change. Moreover, these unchanged things or aspects are becoming increasingly precious and can provide us with a time window. The most important point is that humans are still the dominant force, still telling machines what to do, which is fundamental to maintaining our confidence. This has both technical and institutional implications. That is to say, the two fundamental ways for us humans to cope with job substitution have not undergone fundamental changes so far, although we also need to keep up with the times and constantly correct our direction. Firstly, human capital remains the confidence to withstand the impact of AI, but humans need to know their strengths and weaknesses, and make leveraging strengths and avoiding weaknesses the basic strategy for cultivating human capital in the AI era. So far, human intelligence or natural intelligence still has advantages over AI in the following aspects: (1) soft skills rather than hard skills; (2) Non cognitive ability rather than cognitive ability; (3) Emotional intelligence rather than intelligence; (4) Humanistic understanding and empathy, rather than mathematical and chemical problem-solving abilities, or even coding skills; (5) Implicit knowledge rather than just explicit skills. Secondly, the social welfare system remains a fundamental foundation system, and the material conditions for fulfilling such functions are increasingly strengthened. Marx saw from the development of early capitalism that once labor becomes a commodity, workers find it difficult to escape the fate of exploitation from a systemic perspective. At the beginning of establishing welfare states, Nordic countries emphasized "de commodification" in their institutional design, which weakened the attribute of labor as a purely private factor and strengthened the social rights of workers and their families. In the context of AI's job destruction being greater and faster than job creation, this concept and approach are becoming increasingly important. There are several ways out for workers under the impact of AI employment, whether it is old methods or new ideas. The ways in which employment responds to AI substitution are as follows. Before summarizing these methods, we first provide a reasonable premise that the development of AI will ultimately increase labor productivity by an unprecedented extent. On this basis, workers usually and can have the following ways out. Firstly, transfer to a position that requires higher skills. This is a result that optimistic economists have always believed in, and it has been continuously proven by facts since the occurrence of the Lutheran movement in history. However, this requires workers to have higher skills to adapt to it. In other words, those who obtain this new position and those who lose their old position are usually not from the same group, and to a large extent, they are not from the same queue, or even from the same generation. In the future, the time gap between losing old jobs and obtaining new ones will only become larger. Many economists, including former US Treasury Secretary Summers, have shifted from being confident in technological progress and job creation to now believing that "Lutheranism" has its own reasons. Given that another former US Treasury Secretary, Munuchin, is still "optimistic" about the employment impact of AI, and it is now rare to find someone who holds such an attitude, we can refer to this non naturally occurring position as a "Munuchin style job". Secondly, shift to industries with lower labor productivity and therefore lower returns. The transfer of labor force to higher productivity sectors by Kuznets is a normal direction of industrial structure change, and correspondingly, positions with reduced productivity belong to those with "reverse Kuznets" characteristics. Objectively speaking, the degree of regularization in new positions is lower than in previous jobs. Subjectively speaking, the level of dignity in a new position is also lower than in the previous job. In summary, the quality of employment has been reduced. Thirdly, transfer to a small number of positions in industries with high demand elasticity. This refers to industries where people maintain a huge demand but naturally have the characteristic of difficulty in improving labor productivity. Economist William Baumol regards performing arts as a typical example of this industry. The key to whether this type of industry and position can continue to exist and expand lies in the demand and elasticity of people for corresponding products and services. However, the scale and number of positions in such industries will not expand infinitely. This type of position can be referred to as the Baumol cost disease position. Fourthly, transfer to positions induced by new consumption. Our consumption today may have been unimaginable a few years ago, but even earlier it simply did not exist. The same applies to employment positions. In the future, with the improvement of labor productivity, people's taste will change, new things and concepts will continue to emerge, and as a result, the field of consumption will continue to expand, and the types of professions will be constantly updated. Given that the consumption of such positions is ultimately caused by the improvement of supply side productivity, it is a phenomenon of "supply creating demand", and we can refer to it as "Saai style positions". Fifth, transfer to positions that have emerged due to redefinition. Activities that previously did not meet the definition of employment can now be recognized as "employment" by society and compensated through transfer payments, supported by overall labor productivity. For example, if a person considers themselves a "writer" but has not published any work and received compensation, according to the definition of unemployment surveys, this state of "not engaging in paid work for more than one hour in the past week" is not considered employment. However, if society can afford it, it can also be considered as a form of employment. Similar situations also include those who do not claim to be "working", i.e. those who no longer seek employment. This includes two situations. One way is that the parties involved have a source of support, such as resorting to methods such as gnawing on the elderly to "lie down". Another approach is to receive inclusive social welfare support without the need for employment. For example, if a universal basic income system is implemented, it creates an environment where some people who are impacted by employment choose to no longer participate in traditional jobs. Given that Keynes had long explored ways to share productivity gains, we can also refer to it as a "Keynesian job". Based on the long-term experience of human beings in dealing with the phenomenon of technology replacing employment in economic history, several principled suggestions can be extracted, which are to guide technological development and market behavior through institutional construction, policy adjustments, and institutional reforms, in order to achieve maximum synchronization. Firstly, to maintain the synchronization between the speed of job destruction and the speed of job creation, especially in terms of quantity, to make job transfers as feasible as possible. The second is to maintain the synchronization of productivity improvement speed among various industries and avoid the occurrence of the Solow paradox (IT only improves productivity in some industries but fails to penetrate into other industries). The third is to maintain the synchronization between the speed of AI replacing labor and the speed of training workers' abilities, and to shorten the friction period of re employment as much as possible. The fourth is to maintain the synchronization between productivity improvement and productivity sharing, which is also a requirement and embodiment of the unity of fairness and efficiency. From the perspective of government functions, ensuring the implementation of the above principles can start with several important and urgent response strategies. Firstly, accelerate the construction of a Chinese style welfare state. Several points should be emphasized in this regard: firstly, to accelerate the improvement of the social welfare system with a spirit of seizing every moment, or in other words, at the speed of Moore's Law. Secondly, design and improve the welfare system based on the principle of inclusiveness. This means changing the previous concept of strictly identifying the beneficiaries of social welfare, because in the era of accelerated job loss, it is increasingly difficult to distinguish whether a person is "lying down", and AI expulsion of workers itself has strong externalities. Thirdly, using social solidarity, social protection, and rights protection to offset the spread of informal employment and its adverse effects on workers. Secondly, significantly extend the duration of compulsory or free education. The need to compete with AI places increasing demands on human capital, while also emphasizing the cultivation of non cognitive abilities. A study by the Harvard Center for Child Development shows that in the first few years of life, the brain can establish over one million neuronal connections per second, which cannot be reproduced at any stage of life thereafter. The best time to cultivate non cognitive abilities is between the ages of three and four, and the most ideal measure is to extend compulsory education to this preschool education age. The expected significant increase in labor productivity can significantly expand public education resources and support longer school hours for children. Finally, eliminate institutional barriers in areas such as child development, education and training, mobility and employment, social security, and other basic public services. Especially to eliminate the institutional reasons for the existence of migrant children and left behind children. Research has shown that for the cultivation of children's human capital, especially for the acquisition of non cognitive abilities, as well as their lifelong social mobility opportunities, parental upbringing and care play an irreplaceable role in both school and society. Therefore, promote the reform of the registered residence system and solve the problem of left behind children