Can humans and AI work side by side? A guide to what comes next

Kai-Fu Lee
10 min readNov 25, 2021


A human perspective shift for the 21st Century.

In August, Elon Musk announced that he was developing a humanoid robot called “Tesla Bot” — a prototype of which will supposedly be ready in 2022.

The presentation itself was slightly bizarre, both because the bot was represented on stage by a dancing human in a bodysuit, and because Musk has warned for years about the dangers of artificial intelligence. But he was right about one thing: AI and robotics could eventually lead to a future where technology is so advanced that our role, much of our work and even our purpose in life could fundamentally change.

As someone who has studied and worked in AI for the better part of four decades, my contention is that this change will likely come faster and go further than many of us anticipate — even if it’s driven more by AI-powered software than dancing robots. The question is how we’re going to prepare.

I believe there needs to be a fundamental and historic shift in the way we think about — and relate to — both technology and the world of work. We need a new social contract for an AI-enabled post-work world — and we should start thinking about it now.

The elephant in the room — AI’s upper hand and mass job displacement

Ask any politician what they’re most concerned about, and you’ll usually get the same answer: creating jobs. But by the end of the 21st Century, thanks to AI, we may be thinking about jobs in a completely different way. AI’s main advantage over humans lies in its ability to detect incredibly subtle patterns within large quantities of data — also known as “complex information processing”. This means that intelligent robots are already replacing, in the words of Musk, “dangerous, repetitive and boring tasks.”

It’s easy to imagine warehouse workers, assemblers, construction workers, and plumbers going next. And in the relatively near future, AI will be underwriting our loans, building our homes, assessing our performance and even hiring and firing us — making far more accurate predictions than the best human professionals. But while blue collar workers are most vulnerable to automation, they’re not the only ones.

So what does this mean for each of us and what will this look like on the ground? Robotic process automation or “RPA” is a phrase everyone is going to hear more of in the future. RPA is a “software robot” which can be installed on a routine worker’s computer, and can watch everything the worker does. Over time, by watching millions of people, RPA figures out how to do the routine and repetitive tasks. It confirms with the worker that it has learned a task properly, and later takes over that task, allowing the pool of workers to shrink as their overall workload is lightened.

For example, in a company’s recruitment department of 100 people, RPA can first be applied to resume screening and matching against job descriptions. Let’s say there are 20 people doing that task, and that RPA can help these people be twice as efficient. That’s 10 people who can be displaced. Then, as AI learns from more data, it might replace nearly all of the 20 people.

After that, RPA can take over email communication with the candidates, interview set-up, feedback coordination, hiring decision-making, and even offer negotiation. Each of these steps may displace another 5–10 people. Then AI can do the interview screening, or the first round interviews.

That will save time for HR and hiring managers. All of the above might reduce a total of 90 people for the company. And after recruiting is fully AI-infused, then comes HR training, new employee orientation, and performance evaluation. And after all of HR is AI-infused, finance, legal, sales, marketing, customer service will likely follow. While AI displacement is gradual, eventually it is also nearly total.

As AI does jobs more efficiently, it will generate tremendous economic value. But this shift will also come at a real human cost. In my book, AI Superpowers: China, Silicon Valley, and the New World Order, I estimate that about 40% of our job tasks will be able to be done by AI and automation by 2033. The transition won’t happen all at once, but the trend will be slow, steady, and unavoidable.

While AI may be new, these changes have happened before. In fact, most new technologies were job creators and job destroyers at the same time. The assembly line changed the automotive industry from artisans hand-assembling expensive cars to routine workers building many more cars at much lower prices.

The main difference is that the shift to AI-enabled automation will happen faster. The Industrial Revolution took centuries to spread beyond Europe and the U.S. AI, on the other hand, is being adopted simultaneously all across the world. And with the pandemic accelerating the shift to remote work and challenging the service and hospitality sector, the world is changing even faster.

A fragile fallout

So what will the result of AI displacement look like? To begin with, there will be a growing pool of unemployed workers competing for a shrinking pool of jobs, which will drive down wages. Wealth inequality is likely to go from bad to worse as AI algorithms make some workers redundant while turning their creators into billionaires in record time. If left unchecked, AI in the 21st Century may even bring about a new caste system, with a plutocratic techno-AI elite and everyone else.

Even more problematic than the loss of jobs and income is likely to be the loss of meaning. Ever since the Industrial Revolution, work and meaning have been synonymous for many people. As algorithms and robots outmaneuver humans at tasks they’ve spent a lifetime mastering, the meaning and importance of work and its central place in the lives of millions of workers will be called into question — leading some to question their own worth, and even what it means to be human.

Recent history, including the pandemic, has shown us just how fragile our political institutions and social fabric can be in the face of disruptive change. AI economics could arguably be the largest disruptor in generations — and could lead to geopolitical tumult unlike anything the world has ever seen.

Human-AI symbiosis — unlocking AI’s true potential

No alt text provided for this image
All of this presents a grand challenge to every country and human being on Earth: how can human beings and AI coexist? How can we harness the power of AI without losing our identity as human beings? And what can we — as individuals, companies, and governments — do to prepare for the future we know is coming?

I don’t claim to have all of the answers. But as we continue to understand what AI can do better than humans, it’s important to focus on what AI cannot do as well. Based on my own research and experience, there are a few major areas where AI currently falls short and is likely to struggle for decades to come:

  1. Creativity — AI cannot create, conceptualize, or plan strategically. While AI is great at optimizing for a narrow objective, it is unable to choose its own goals or to think creatively. Nor can AI take a certain area of knowledge and apply this across domains or use common sense.
  2. Empathy — AI cannot feel or interact with emotions like empathy and compassion. Therefore, AI cannot make another person feel understood and cared for. Nor can it motivate. Even if AI improves in this area, it is unlikely that humans would opt for interacting with an apathetic robot for many human-touch services.
  3. Physical dexterity — AI cannot accomplish complex physical work that requires dexterity or precise hand-eye coordination, or deal with unknown and unstructured spaces, especially ones that it hasn’t observed.
    This means that, while there are categories of jobs that are at risk of becoming obsolete thanks to AI, it isn’t all bad news.

There are categories of jobs that are likely safe — at least for the time being. As was the case in the Industrial Revolution, there will be new jobs where humans and new AI tech can work together and complement each other. There will also be jobs where AI will be used as a tool to enhance human creativity, productivity, and dexterity. A robot may mix better drinks, for example, but a bartender possesses better social skills.

Preparing for the AI economy

This is why, as we reimagine living and working in the age of AI, I propose focusing on the three R’s — relearn, recalibrate and renaissance.


The first step is for us to relearn. This is the process of helping people displaced from unskilled, routine, and quantitative jobs to learn new skills. People in endangered jobs should be made aware of their situation and the need to adapt. Vocational schools need to redesign their curriculum to reduce courses for doomed jobs, and increase courses for sustainable jobs.

Governments could take the lead and provide incentives and subsidies for those courses, rather than blindly pursuing broad-brush economic measures like universal basic income.

Corporations should follow the lead of Amazon’s Career Choice program that pays up to $12,000 annually over four years for Amazon’s hourly employees to earn degrees in high-demand occupations.

The importance and number of human-centric service jobs, such as nursing, will grow as wealth and life spans increase. The World Health Organization predicts that by 2030, we will fall short of the number of healthcare workers required to achieve the United Nations Sustainable Development Goal of “good health and well-being for all” by approximately 18 million. There is an acute need to reassess such vital, yet undervalued, human-centric service roles both in terms of how they are perceived and how much they are paid. These jobs will form a bedrock for the new A.I. economy.

Another way to balance the supply and demand of jobs is to turn today’s volunteer services into important full-time jobs in the future. During this era of automation, for example, there will be tremendous demand for volunteers to answer hotlines for displaced workers, dealing with their questions and anxiety. These volunteers could potentially become eligible for extra UBI payment, and able to do it full-time.

In my new book AI 2041 I imagine many such scenarios. One story explores “job re-allocators,” a new kind of company that is funded by the government and looks after people being displaced. Job re-allocators will train AI-displaced workers to gain new skill sets in job categories that fit each worker’s skill set and preference.

The AI job displacement tidal wave will eventually take away virtually all routine jobs, which tend to be entry-level jobs. But if no human takes an entry-level job, how do they learn, grow, and advance to more senior and less routine jobs? As automation becomes pervasive, we will also need to make sure there are still ways for people to enter all professions, to learn by doing, and get promoted based on their new capabilities.

Another scenario pursues the possibility of entry-level workers doing “virtual jobs” that have little economic value but help develop their skills, advance their careers, and realize their potential. AI and AR technologies can play a significant role to help employ these entry-level jobs and facilitate their careers, as well as provide a sense of worth, contribution, and even self-actualization.


Second, as we prepare people for the eventual transition to an AI-driven economy, we will also need to recalibrate many of today’s jobs. Like software did a few decades ago, A.I. can augment a human’s creative thinking with a computer’s relentless ability to churn through masses of data, hypothesize alternatives or optimize outcomes. There will not be a single, generic AI tool, but specific tools customized for each profession and application. We may have an AI-based molecule generation program for drug researchers, an AI advertising planner for marketers or an AI fact-checker for journalists.

Merging AI optimization and the human touch will reinvent many jobs and create even more. An ideal symbiosis would be where AI takes care of routine optimization tasks in tandem with humans, who will carry out tasks that require the human touch of warmth and compassion. For example, the future doctor will still be the primary point of contact trusted by the patient, but will rely on AI diagnostic tools to determine the best treatment. This will redirect the doctor’s job into more of a compassionate caregiver, whose education can be shortened, thereby reducing the cost of healthcare, and giving patients more time with their doctors.

The coming of A.I. will create jobs we cannot even conceive of yet. These will include advanced jobs like programming AI and robots, professional jobs like robot and autonomous vehicle repair, and also entry-level jobs like data processing and labeling — to create more “food for AI”. AI will create many other types of jobs that we cannot yet predict, just as the Internet has over the past two or three decades (for example, roles like the Uber driver). We should watch for the emergence of such roles, make people aware of them and provide training for them.


Finally, just as the wealthy Italian cities and merchants funded that country’s Renaissance, AI should fund and inspire a renaissance of its own.

With machines taking over many duties and tasks in the new economy, AI will inject flexibility into traditional working patterns, allowing us to rethink what work-life balance should look like and transforming both the weekday routine and retirement thresholds. With more freedom and time in such a new social contract, people will be liberated to follow their passions, creativity and talents, and to let that personal exploration inform their careers as never before.

All of us have a role to play in making this a reality. Companies will need to retrain a massive number of displaced workers. Schools need to reinvent education to produce creative, social, and multi-disciplinary graduates. Governments must raise an astronomical amount of money and redistribute it to fund this transition. It is crucial to redefine the work ethic of society, entitlements for citizens, responsibilities for corporations, and the role of governments. In short, we will need to rewrite our social contract — and we should start now.

The rise of AI has serious implications for society and the world of work. But if we can successfully adapt towards symbiosis, I believe that AI will liberate us from the mechanical drudgery of routine tasks and allow us to focus on our humanity and the compassionate connections between us all. We should get to work.

You can grab a copy of AI 2041: Ten Visions For Our Future at the following link: AI 2041: Ten Visions For Our Future. Available now.

The article was first published on LinkedIn on September 15, 2021



Kai-Fu Lee

AI Expert, CEO of Sinovation Ventures (创新工场), founding President of Google China, Author of “AI Superpowers”