How many jobs will GPT destroy for migrant workers?

Today we will continue to talk about
the impact of GPT on work. First, I will share a dark method of using AI
that I recently saw on the Internet. Next, in many companies that are indifferent to AI, especially non-technical companies, it may actually be the next year or two. The golden age of working as a dawdler because the subordinates who understand AI can do their work better than before in 1/10 of the time , but the boss who does not understand AI will give them promotion and salary increase. Do you know what this means? It means that it will become a new game in the workplace to let the boss understand the capabilities of AI as slowly as possible while allowing oneself to understand the capabilities of AI as quickly as possible . What about human feelings , but human beings have already begun to use human feelings and sophistication, so human beings have the creativity of reverse thinking, but this kind of opportunism can be confused for a while, but not forever . They will disappear in large numbers, so what are work tasks? For example, writing a promotional copy for a product, drawing an illustration for a PPT, making an execution schedule for a project , or writing a piece of front-end code for a website. These are all tasks.

































And the biggest problem in most of our traditional companies is that too many jobs
are created that are only responsible for
completing tasks ? Screw? So this kind of weird dislocation will be completely destroyed sooner or later in the coming AI era It is the penetration rate. Next week, we will use a special issue to talk about this paper , so today we can talk about the general direction. In fact, I quoted a McKinsey research report two years ago , and they found that white-collar jobs today In the content , creation and decision-making time accounted for less than 7%, professional skills contribution time accounted for less than 14% , communication time accounted for about 16%, accident handling time accounted for 12% , and all these time accounted for about half . The other half of the time is spent on information collection and information processing , as well as repetitive, process-based and automated work content. So if we look at this work ratio, we will find that































First of all, the blue part
, information collection and processing
, and things related to repetitive work content
will be replaced by AI in the future.
In the orange and green parts,
the so-called professional skills module
will also be greatly challenged.
For example,
the current machine Tens of thousands of lines of code can be written in half an hour
, so those codes that were not dared to rewrite
before are dared to write now
, so that the system update may be accelerated
, and the group of experts who are familiar with the old system will easily fall behind
if they cannot continuously update their knowledge
For
another example, painters with strong hand-painting skills may be unwilling to use AI software to work
because of their own abilities. Then, tomorrow, when AI drawing skills are getting stronger and stronger, most of the professional skills they used to be proud of may also change . Therefore, the job itself will not disappear in the future, but the whole society may redefine a large number of jobs , and for those jobs that have been redefined, the proportion of different modules may also change as follows : In working hours, the creation and decision-making time, together with the learning and research time, will drop to more than 50% , and then communication and collaboration and dealing with accidents may account for another 40% and finally 10% of the time

















People will use it to tell AI software what to do
, and most people don’t need to be able to use PS or write fans to make the machine work well. They
use
natural language programming
, which is the most popular Prom engineering now.
In other words,
we are directly using the language of human-to-human communication
to guide the machine,
help us make websites, cut videos, write articles,
generate research reports
, and even conduct user interviews automatically.
Hey, this means that
one person can suddenly complete
the tasks of dozens of people.
The key to the problem It’s not here yet. The crux of the problem is what kind of ultimate goal is achieved by
completing so many tasks ? In other words, research on how to achieve the goal is what most of the future work needs to do. For example, what we want is not a nail on the wall . What I want is to be able to hang things , but most of the work in the past was actually arranged to drive nails on the wall. What needs to be studied in the future work is who is the owner who wants to hang the nails and what is his budget? He needs to hang them on this wall. What kind of things is the upper limit of the quality of the things he hangs ? Can this wall be destroyed at will? The style of the wall, what kind of hangings should be matched to be more beautiful , etc.


















In the eyes of the person who is responsible for nailing the nails,
it is inexplicable
, but in the future,
these seemingly inexplicable problems in the past
are actually the key to most of the work. In fact , I found that Seven Stars is the most comprehensive
in terms of new jobs in the AI ​​​​era and
what new capabilities are required.
One of the statistics is that McCann 7 listed 56 skills in four categories and 13 groups
in a report before. I will not list the 56 skills one by one. Interested students can take screenshots . Anyway, it doesn’t matter whether he lists the 4 categories of skills. Whether it is the cognitive ability of reasoning , flexibility, and overall planning, or the numerical ability of understanding the working principle of data and algorithm AI , or the commonplace interpersonal communication ability, in fact, they all fall into the category of comprehensive numerical value that we usually think of , but there is only one It is self leadership, which sounds like an amazing ability. In our past education system, it should have been an elective item , but in the future where AI assistants are extremely powerful, it may become a compulsory course for most people because no matter how powerful AI is. He also needs a master. You see, even if there are thousands of troops serving you , they also need a king who can point out the direction. Unfortunately, our entire education system and jobs are designed now.
























It is impossible to train qualified kings
well. There are even many people who
actually instinctively dislike becoming a king.
Because in the past,
becoming a king meant leading many people, and there
was a lot of pressure.
But in the future, one person is an army.
So the meaning of becoming a king will be completely different.
Speaking of this, I
will talk about the impact of GPT4 on education. I firmly believe that this impact will become very obvious
in the next few years. As you know, as a 42-year-old middle-aged person Almost all of my peers are parents of teenage children , and the emergence of GPT4 has given us a huge sense of crisis in China’s primary and secondary education and the entire teaching system . So I also recently opened a new account . After discussing the arrival of GPT, we urgently need to create various patches for the current education system. Although the current education system has been working very well in the past 200 years when human intelligence has killed machines, it has been in just a few months in the past 7 years. So we're suddenly entering a new era where machine intelligence surpasses human intelligence, and that's obviously what makes education work for the past 200 years.
























Suddenly it stopped working
, so what kind of new education method
will be useful in the next 10 years?
I am very curious about this question.
Although I am not an education expert
, my parents
have also been middle school teachers all their lives
, so I have been very
close After getting in touch with this industry,
I also found that in fact,
many good methods have emerged around the world,
but
the vast majority of Chinese parents may not understand them,
such as personalized education, individually configured education,
project-based education, project-based learning,
social-emotional education, search for emotional learning
, and China. The emerging interdisciplinary education
contains many brand-new educational concepts
that can help current children become beneficiaries rather than victims
in the age of AI . Therefore, although we cannot change China's current education system in a short time , we can at least see its future clearly. Insufficient, I will actively patch him. Well, that's all for this article. Thank you for reading! !