Introduction: The People With Hammers
On a winter night in 1811, shadows descended on a textile mill in Nottinghamshire, England. They carried heavy hammers, and they had a single target: the newly installed knitting machines. The sound of machinery being smashed to pieces rang out in the darkness. These people called themselves Luddites, after a legendary figure named Ned Ludd.
Today we use the word Luddite as a sneering label for anyone who reflexively hates new technology. But the circumstances of those hammer-wielding people more than two hundred years ago were far from simple. They did not hate technology itself. What they could not bear was a change that rendered their hard-won skills and their livelihoods useless in an instant.
We now stand before a similar question. As artificial intelligence advances rapidly, especially generative AI that writes text, draws pictures, and produces code, many people ask: Is my job safe? This essay does not try to answer that question in one stroke. Instead, by taking a long look back through history and placing many perspectives side by side, it aims to help us ask better questions.
So, let us set the hammer down for a moment and think together.
The Lump of Labour Fallacy: Is the Pie of Jobs Fixed
First, we must address one of the most common misconceptions: the lump of labour fallacy.
This fallacy assumes the following. The amount of work that exists in the world is fixed. Therefore, when a machine takes one job, the work available to people disappears forever by that same amount. At first glance it sounds obvious. Yet economists have long pointed out that this assumption is wrong.
Why is it wrong? Because the amount of work is not a fixed pie. New technology eliminates existing work while at the same time creating new work and entire industries that did not exist before. The automobile reduced the work of coachmen who drove carriages, but it created new occupations: auto mechanics, gas station attendants, road construction workers, and car insurance agents.
Consider this. If you told someone a hundred years ago, Your grandchildren will be search engine optimization specialists, data scientists, and YouTube creators, they would have looked at you as if you were speaking an alien language. The belief that the total amount of work is fixed overlooks how hard it is for us to imagine in advance what new kinds of work the future will bring.
There is one caution, however. Pointing out the lump of labour fallacy does not let us leap straight to the conclusion that there is nothing to worry about. The fact that new jobs emerge in the long run is a completely different matter from whether a particular generation or a particular occupation avoids great suffering during the transition. We will return to this point later.
What History Tells Us
The Industrial Revolution: An Age of Fear and Abundance Together
The Industrial Revolution, which began in late eighteenth-century Britain, was the greatest transformation of labor in human history. Steam engines and weaving machines took away the work of countless cottage artisans who had spun thread and woven cloth by hand. The Luddite movement mentioned earlier arose precisely during this period.
In the short term, this change was brutal. The wages of skilled weavers collapsed, and many families were driven into the harsh conditions of factory labor. Child labor, long working hours, and dangerous workshops were the dark shadow of this era.
In the long run, however, the Industrial Revolution explosively raised humanitys material abundance. According to research by economic historians, per capita income, which had been almost stagnant for hundreds of years before the Industrial Revolution, began to rise steadily over the course of the nineteenth century. Much of the abundance we enjoy today is the fruit of the change that began then.
Here lies an important lesson. The short-term pain and the long-term benefit of technological change are not distributed equally to the same people. It was mainly the next generation that picked the fruits of abundance, while it was the generation standing in the middle of the change that endured the pain.
ATMs and Bank Tellers: The Paradox of Automation
Let us look at a more recent example, one that runs counter to intuition: the story of the automated teller machine, the ATM, and the bank teller.
Common sense would say this. When ATMs spread, machines take over the deposit and withdrawal work that people used to do, so the number of tellers should fall. Yet according to the analysis of economist James Bessen, who studied the actual case in the United States, something interesting happened.
As ATMs spread, the number of tellers needed to run a single branch did indeed fall. But as the cost of operating a branch dropped, banks opened more branches. As a result, the total number of tellers actually rose for a time. Moreover, the role of the teller changed. The center of gravity shifted from simply counting money toward recommending products to customers and managing relationships.
The lesson of this case is subtle. Automation does not necessarily reduce the total number of jobs; sometimes it changes the nature of the work. When machines take over the repetitive and simple parts, people often come to focus on more complex, relationship-centered work. Still, this case is not a universal law either. In other industries, automation has indeed sharply reduced the number of jobs.
The Printing Press and Scribes: The Twilight of the Monastic Scriptorium
Let us go back much further than the ATM. Until Johannes Gutenberg made movable-type printing practical around 1450, making a single book was an enormous labor. European monasteries had spaces called scriptoria, where copyist monks transcribed text by hand, one letter at a time. Completing a single thick book took months, sometimes years. Books were that rare, that expensive, and the preserve of a privileged few.
When printing spread, this scene changed rapidly. A printing press could turn out in a few days what one person might copy in a lifetime. Demand for professional copyists who transcribed by hand fell sharply. The skilled craft of fine handwriting lost its value, at least when it came to producing books in large numbers. In this respect, the scribes could be called distant ancestors of the sense of loss the Luddites would feel some two hundred years later.
But the ending of the story was not simple disappearance. As the price of books fell, the number of readers exploded, and as a result new work and industries arose in a chain: authors who wrote, publishers who printed, typesetters who arranged the type, bookshops that sold, and papermakers who made the paper. As knowledge spread more widely, society as a whole was shaken, from education and science to the Reformation. In other words, in exchange for eliminating a handful of copying jobs, the printing press created new labor and value on a vastly larger scale.
Here too the pattern we saw earlier repeats. The person who suffered in the place of the vanished occupation and the person who enjoyed the newly opened abundance were not necessarily the same. An elderly scribe who had honed his handwriting over a lifetime could hardly become a typesetter or publisher overnight. Technological transition almost always carries this kind of mismatch.
The Spreadsheet Revolution: From Bookkeeping Clerk to Analyst
There is one more example, closer to our time and remarkably clear: the computer spreadsheet. In 1979, Dan Bricklin and Bob Frankston released a program called VisiCalc. You entered numbers into a grid on the screen and set formulas, and the moment you changed the value in one cell, every connected cell recalculated automatically. In the 1980s, Lotus 1-2-3 followed and became the office standard, and later Microsoft Excel carried this current forward.
Before the spreadsheet appeared, the finance department of a large company had countless clerks organizing ledgers by hand, adding up numbers and checking the sums. For a company to find out what would happen if revenue rose by ten percent, a person had to recompute every number from scratch on a paper ledger. It was grueling work that took days.
The spreadsheet took over that repetitive calculating labor in an instant. So did all accounting-related jobs disappear? Interestingly, no. Jobs that simply added numbers and checked sums clearly declined. But at the same time, the value of interpreting and using numbers grew. As calculation became easy, companies tried countless what-if scenarios they had never dared attempt before. The role of the accountant shifted from someone who organized ledgers toward an analyst who reads what the numbers mean and helps make decisions.
This case shares a texture with the ATM story above. When automation takes over a task, peoples work often moves to a higher level of judgment and interpretation. But here too there is a caution. Those who passed smoothly through this transition were the people who quickly learned the new tool and built a new capability called analysis beyond simple calculation. For those who did not, the same change arrived as a threat. The fact that even with the same tool the outcome differed from person to person is the important clue the spreadsheet revolution left us.
The Telephone Operator: Portrait of a Vanished Job
There is also a counterexample. In the early twentieth century in the United States, the telephone operator was a stable job for many young women. In those days, when you placed a call, an operator connected the line by hand. But as automatic telephone exchanges spread, this occupation almost entirely disappeared.
In this case, automation did not change the nature of the job; it eliminated the occupation wholesale. Of course, as the telephone network grew, new jobs arose in other fields, but the very people who had sat at the switchboard did not move smoothly into those new jobs.
Placing these two cases, the ATM and the telephone operator, side by side, we can see that the result of automation does not flow in only one direction. Some occupations survive by changing their form, and some disappear. Which one a given occupation becomes depends on the nature of the technology, the response of the market, and the choices of society.
A Timeline of Automation and Work
1769 James Watt patents his improved steam engine — accelerating the Industrial Revolution
1811 The Luddite movement begins (Nottinghamshire, England)
1870s The Second Industrial Revolution — the age of electricity and mass production
1913 Ford introduces the moving assembly line
1920s Automatic telephone exchanges spread — telephone operator jobs decline
1950s The arrival of the computer; the word automation enters common use
1970s ATMs begin to be introduced
1990s The internet and the PC go mainstream — office work transforms
2010s The rise of smartphones, the cloud, and machine learning
2022 Generative AI based on large language models spreads to the public
2020s Generative AI expands into the domain of knowledge work
Looking at this timeline, one pattern stands out. Every time a new technology appeared, people cried out that this time is different. And in fact, it was a little different each time. So is this wave of AI really, truly different?
Is It Really Different This Time: The Singularity of Generative AI
Past automation mainly targeted physical labor and repetitive tasks. Machines were good at lifting heavier loads than people and repeating the same motion without tiring. By contrast, creative work, writing, drawing, and devising strategy, was long regarded as the unique domain of humans.
Yet generative AI has set foot in precisely that domain. It writes text, draws pictures, composes music, and writes code. This is why many people feel that this time is different. For the first time, the blade of automation is pointed at knowledge work and creative work.
But there is much to weigh carefully. First, todays AI quickly produces plausible-looking output, but it still shows limits in factual accuracy, deep contextual understanding, and responsible judgment. Second, the fact that AI automates part of a task does not mean the whole occupation disappears. An occupation is usually a bundle of countless small tasks.
This second perspective leads directly into our next topic, the debate over replacement versus augmentation.
Replacement or Augmentation: Two Futures
When we talk about the impact of AI on work, two broad scenarios stand opposed.
Scenario 1: Replacement
The first is the view that AI replaces people. The scenario holds that once AI becomes smart enough, companies will use AI instead of people to save costs, and as a result many jobs will disappear. There is particular concern that work with relatively clear patterns, such as formulaic writing, simple translation, basic customer service, and some coding tasks, will be affected first.
Those who support this view say: Past machines replaced hands and feet, but this AI replaces the mind. If work that uses the mind disappears, there is no obvious place for people to move to.
Scenario 2: Augmentation
The second is the view that AI augments people, that is, makes them stronger. AI is not a replacement for people but a tool that extends human ability, just as the calculator did not eliminate mathematicians but instead led them to tackle more complex problems.
From this perspective, AI takes on the dull and repetitive parts while people focus on the parts that require judgment, creativity, empathy, and responsibility. Doctors, supported by AI in diagnosis, talk more deeply with patients; lawyers craft more refined strategies on the basis of material AI has organized; writers explore new expression by exchanging ideas with AI.
Look at Tasks, Not Jobs
There is one perspective that makes the replacement-versus-augmentation debate considerably clearer: looking at the unit of the task rather than the unit of the job.
What we commonly call a single job is in fact a bundle of dozens of small tasks. The day of one accountant, for example, consists of many tasks of differing character: data entry, checking sums, writing reports, meetings with colleagues, consulting with clients, and judging anomalies. The fact that AI skillfully handles some of these tasks, such as data entry and checking sums, does not mean the whole occupation disappears. Rather, to the extent AI relieves the simple tasks, people can spend their time on harder tasks such as judgment and consultation.
This task-level perspective tells us two things at once. On one hand, almost every occupation will be affected by AI to some degree. Occupations completely untouched are rare. On the other hand, that impact does not immediately mean the occupation will vanish. In most occupations what gets automated is some of the tasks, not the whole job. So the more accurate question is not Will my job disappear but Among the tasks that make up my job, which will be automated, and so where will I come to focus more.
Reality Lies Somewhere in Between
The actual future probably lies somewhere between these two scenarios. The same technology will act mainly as replacement in some roles and as augmentation in others. And which prevails depends not only on the pace of technological advance but greatly on the choices of companies, the adaptation of workers, and the institutions of society.
There is one thing to remember here. Replacement or augmentation is not a fate unilaterally determined by technology; it is closer to a choice that we shape together.
Changing Jobs, Unchanging Value
Work Strongly Affected and Work Less Affected
Not all work is affected equally. In general, the following tendencies are observed.
Work relatively more affected
- Formulaic text writing (report drafts, standardized documents)
- Simple data entry and organization
- Basic translation and summarization
- Customer service with clear patterns
- Repetitive code writing
Work relatively less affected
- Complex physical work (plumbing, electrical work, caregiving)
- Work requiring deep interpersonal relationships and trust
- Work involving high responsibility and ethical judgment
- Non-routine, creative strategy-making
- On-the-spot improvised problem solving
What is interesting is that in the age of AI, the value of skilled manual labor is being reappraised. Unclogging a blocked pipe, fixing broken wiring, and caring for people directly are hard to replicate digitally. Work once undervalued as simple labor may instead become the solid ground that withstands the wave of automation.
Comparison Table: The Industrial Revolution and the AI Revolution
| Comparison | Industrial Revolution | AI Revolution |
| --- | --- | --- |
| Main target of replacement | Physical labor, handicraft | Knowledge work, some creative work |
| Pace of change | Gradual over decades | Relatively fast |
| Emergence of new jobs | Factory labor, machine maintenance | Data science, AI management roles |
| Shift in required skills | From hand skill to machine operation | From memorization to judgment and collaboration |
| Social response | Labor law, expansion of public education | Reskilling, policy debate underway |
This table is only a simplified comparison and cannot capture every detail. But it shows at a glance how the two revolutions are alike and different. The key is the shift in required skills. If the Industrial Revolution demanded a transition from hand skill to machine operation, the AI Revolution is demanding a transition from simple memorization and repetition to judgment, creativity, and collaboration.
One thing to add: it is best not to brush past the pace of change row in the table. The Industrial Revolution unfolded over decades, even a century, so generations born during it could naturally adapt to new work as they grew up. By contrast, if the pace of the AI Revolution truly is faster, a situation may arrive in which the same person must change occupations several times within a single lifetime. This is a new kind of challenge that was rare during the Industrial Revolution.
New Jobs and the Skills Gap
Technology does not only eliminate jobs. It also creates new ones. Think of occupations that did not exist just twenty years ago: app developer, data scientist, social media manager, cloud engineer, content creator. And now the age of AI is creating still other new occupations: people who refine AI models, people who review and take responsibility for AI output, people who design collaboration between humans and AI.
But here lies a trap called the skills gap. Even if the newly created jobs make up for the vanished jobs, if the two jobs require completely different skills, the person who lost work cannot immediately obtain new work.
A telephone operator cannot become a data scientist overnight. Bridging this gap requires time, money, and above all the opportunity to learn. And this opportunity is not given equally across people, regions, and classes. It is precisely here that technological change meets the problem of inequality.
The Policy Debate: What Should We Do
How should society handle the shock that technological change causes? Many strands of opinion contend over this question. This essay does not advocate for any one side; it tries to present the major positions as fairly as possible.
The Universal Basic Income Debate
One of the hottest debates is universal basic income, or UBI: the idea of regularly paying every citizen a fixed amount with no conditions.
The logic of those in favor runs like this. If AI automates many jobs, the existing income structure that depends solely on employment may be shaken. A basic income provides people with a minimum safety net, cushioning the shock of sudden change and giving people room to attempt new things such as reskilling, starting a business, or caregiving work.
The logic of those opposed is far from negligible. First, paying money to every citizen requires enormous funding, and it is unclear how that funding would be raised. Second, there is concern that if income is guaranteed even without working, peoples motivation to work may decline. Third, some argue it would be more efficient to concentrate that funding on those who truly need help.
In fact, small-scale basic income experiments have been carried out in various places around the world, and their results are interpreted in conflicting ways. Some studies report that basic income was positive for life stability and mental well-being, while others find that its effect on working hours or employment was limited or mixed. In short, it is still too early to draw a clear conclusion.
Other Policy Options
Basic income is not the only alternative. Several other approaches are discussed alongside it.
Major policy options (not mutually exclusive)
1. Support for reskilling and lifelong learning
- The state helps people who lost work learn new skills
2. Strengthening the social safety net
- Reinforce existing systems such as unemployment benefits and job transition support
3. Universal basic income
- Pay everyone an unconditional basic income
4. Reducing working hours
- Share the gains of higher productivity through shorter working time
5. Taxing technology and automation
- Return part of the gains from automation to society
Each option has its own strengths and limits, entangled with value judgments about who bears the cost and to whom the benefit returns. Some societies will mix several options, and others will choose a different path. There is no single right answer; this is ultimately a matter of that societys values and consensus.
Different Responses by Country and Sector
Even when facing the same wave of automation, the ways societies respond differ markedly. Rather than declaring which side is right, it helps to examine which values each different approach prioritizes.
At one end is the reskilling-heavy approach. Some Nordic countries are often cited as examples. When a worker loses a job, they place weight not on holding that person in a single workplace but on helping with the transition of the occupation itself. They support living through generous unemployment benefits while at the same time helping people move into new jobs through active job training and reskilling programs. This model, often called flexicurity, is close to the idea that layoffs are relatively free but the shock is borne by society together. Its strength is that social resistance to change falls and the labor market becomes flexible; its limit is that it requires the foundations of enormous funding, high taxes, and social trust.
At the other end is the market-led approach. The United States is often mentioned as an example. This approach minimizes government intervention and is closer to leaving the labor market to find a new equilibrium on its own. Companies hire and fire relatively easily, and new industries arise and vanish quickly. Its strength is that the pace of innovation and new job creation can be fast; its limit is that the shock of transition can fall more heavily on the individual, and gaps tend to widen in the meantime.
Sectors differ in texture too. Manufacturing introduced robots and automation equipment early on, and in the process the number and nature of jobs changed greatly. By contrast, sectors such as caregiving, education, and healthcare, where relationships and trust between people are central, automate relatively slowly. And even within the same country, the size of the shock differs for large cities versus the provinces, and for large corporations versus small businesses.
The important point here is that no model is a perfect answer. The reskilling-heavy approach values stability and fairness but costs a great deal; the market-led approach values dynamism and efficiency but the distribution of the shock may be uneven. Each society chooses somewhere in between according to what it values more. And there is no answer key for that choice.
A Thought Experiment: The Automated Town
Let us move the abstract discussion into a concrete scene for a moment. Imagine a fictional town.
Suppose there is a small town of twenty thousand people. A considerable share of the towns jobs are concentrated in a nearby logistics warehouse, a call center, and an office that handles accounting work on contract. Then, as new automation technology and AI are introduced rapidly, in just five years forty percent of the towns jobs disappear or change greatly. Robots take over the warehouses sorting work, AI takes over the call centers simple service, and software takes over the accounting offices repetitive work.
So what unfolds before this town? The first thing to strike is the shock. When many people lose work all at once, their spending falls along with it. Shops that counted those people as customers, like restaurants, hair salons, and neighborhood stores, see their revenue drop too. Even people who did not directly lose work to automation are hit in a chain reaction. The towns tax revenue falls, and it becomes hard to maintain public services such as schools and libraries.
So what choices can this town make? There are several paths.
The options facing the automated town
1. Leave — move to another city in search of work
(but not everyone can leave, and the town empties further)
2. Reskill — learn new skills and move into different work
(but who pays the cost, and what jobs are waiting)
3. Attract new industry — bring in new work suited to the age of automation
(but such opportunities do not come evenly to every town)
4. Hold out — cling to the old ways and delay change
(but over time, falling behind in competition is likely)
5. Share together — return part of the gains of automation to the community
(but those gains often go to companies outside the town)
What this thought experiment shows is that the shock of automation is not merely a matter of the number of jobs. It is a matter that shakes a communitys spending, tax revenue, population, public services, and even peoples self-respect all together. One more thing is clear: the faster and more concentrated the shock, the harder it is for society to absorb. Even the same forty percent is a completely different story depending on whether it happens slowly over fifty years or strikes in just five.
What divides this towns fate is not the technology itself. What decides the ending is what preparation the town and the society beyond it have made, and who shares the shock of change and how. And this is a question that applies just the same to the real world we live in.
Why Work Means More Than Money
So far we have mostly talked about jobs as a source of income. But the meaning that work holds for a person is far broader than that. If we leave this out when talking about automation and AI, we see only half the problem.
First, work is a source of meaning and purpose. Many people gain through the work they do a sense that I am doing something useful for the world. The experience of making something, being of help to someone, and contributing to some outcome gives a person a reason to live. Even if income is guaranteed unchanged, when this sense of meaning vanishes, a person often feels an emptiness.
Second, work creates the structure of time. Waking in the morning and going somewhere, having a rhythm to the day, building life in units of a week, much of this whole framework comes from work. When work suddenly disappears, it is not only money that disappears but also the familiar framework for how to fill a day that is shaken along with it.
Third, work is a site of social belonging. The workplace is an important stage where people meet colleagues, accomplish something together, and gain the feeling of belonging to a community. To lose work often means being cut off from that social network. This severance can lead to loneliness and isolation.
Fourth, work is deeply entangled with identity and status. One of the questions we most commonly ask someone we have just met is, What do you do? Like it or not, in many societies people explain and are judged by their occupation. So the loss of work can go beyond mere economic loss into a shaking of the identity question of who am I.
Why do these four dimensions matter? Because even if we assume that in the future, with automation advanced enough, the money problem is solved for everyone in some form such as basic income, the other functions of work, meaning, structure, belonging, and identity, are not automatically filled. When we ponder the future of work, the question we must ask is not only how will people make a living. Where will people find meaning and belonging is just as heavy a question. This is not a matter of medical diagnosis but a matter of society and culture that we must design together.
A Quick Quiz: Test Your Intuition
Think about whether each statement below is true or false. The answers are right beneath.
Question 1. Right after ATMs were introduced, the number of bank tellers in the United States immediately fell sharply.
Question 2. The lump of labour fallacy refers to the mistaken assumption that the number of jobs in the world is fixed.
Question 3. Universal basic income is a policy that all economists support and that is beyond dispute.
Question 4. Spreadsheet programs such as VisiCalc, which appeared in 1979, made all accounting-related jobs disappear.
Question 5. As Gutenbergs printing press spread, demand for scribes who copied books by hand fell sharply.
Question 6. Even when the same proportion of jobs is automated, the shock to society differs depending on whether the change happens over fifty years or strikes in just five.
Now let us check the answers.
Answer 1. False. As we saw earlier, even after ATMs were introduced, the number of tellers in the United States actually rose for a time as branches increased. It was a result contrary to intuition.
Answer 2. True. The lump of labour fallacy refers to the misconception of seeing the amount of work as a fixed pie. Because new technology also creates new work, this assumption is often wrong.
Answer 3. False. Universal basic income is a representative debate topic on which opinion is sharply divided. Views differ greatly over funding, work motivation, efficiency, and more.
Answer 4. False. The spreadsheet did reduce some jobs, such as simple calculation and checking sums, but it did not eliminate all accounting-related jobs. Rather, the nature of the work changed as the value of the analysis role, interpreting numbers and helping make decisions, grew.
Answer 5. True. As printing spread, demand for scribes who copied books by hand fell sharply. In its place, however, new work and industries arose, such as authors, publishers, typesetters, bookshops, and papermakers.
Answer 6. True. Even with the same proportion, the pace of change matters greatly. The faster and more concentrated the shock, the harder it is for society to absorb and adapt to it.
If you got all six questions right, you already have a good grasp of the heart of this topic.
Pace and Distribution: The Two Real Variables
Stepping back to organize the discussion so far, the key variables that decide the future of work narrow down to roughly two: the pace of change and the distribution of benefit and burden.
First, pace. As we saw in the automated town earlier, even a change of the same size brings completely different suffering to society depending on whether it comes fast or slowly. When change happens slowly over a generation, people buy time to adapt through natural retirement and new entry. As the old scribe nears retirement, the young generation steps into the printing industry from the start. But when change rushes in over just a few years, this buffer cannot work. Much of the current anxiety surrounding AI comes not only from what AI can do but from how fast it gets there.
Next, distribution. Even if the total abundance that technology creates increases, to whom that abundance returns is a separate matter. The benefit of automation tends to concentrate in the people and companies that own the technology, while its cost tends to concentrate in the workers who lost their jobs. The reason abundance spread evenly only to the next generation during the Industrial Revolution is that the distribution did not happen on its own; it became possible only through the choices of labor law and social institutions.
The reason these two variables matter is that both are areas where people can have influence, not areas unilaterally set by technology. We can design institutions that regulate the pace of change, and we can decide together how to divide benefit and burden. This is why the debate surrounding AI ultimately returns to a question not of technology but of societys choices.
Interestingly, pace and distribution are entangled with each other. When change is too fast, the shock strikes before institutions to adjust distribution can even be prepared; when distribution is too lopsided, social resistance demanding that the pace be slowed grows fierce. So rather than treating these two variables separately, a perspective that views them together is needed. The question of how to receive a technology is, in the end, the task of solving both how fast and for whom at the same time.
Strategies for Adaptation: What Can the Individual Do
In the face of vast change, the individual easily feels powerless. Yet history shows that people and societies who adapted to change certainly existed. Let us organize a few directions. These are not the right answers but starting points for thought.
1. Do Not Fear the Tool; Tame It
The heart of the Luddites who smashed machines with hammers is understandable, but in the end it did not stop the age of the machine. The same is highly likely for AI. So rather than rejecting it outright, it is more realistic to understand the tool and learn how to use it in your own work. The gap between those who use AI well and those who do not may become an important dividing point going forward.
2. Invest in Abilities Machines Are Bad At
It helps to distinguish between work AI does quickly and skillfully and work that people still do better. Deep contextual understanding, ethical judgment, genuine empathy, creative problem solving in complex situations, building trust between people. Such abilities are hard to automate in the short term.
3. Cultivate the Ability to Learn Itself
The era of living off a single skill for a lifetime is fading. What matters more is the ability to learn new things quickly. The faster the change, the more an attitude of continual learning becomes a greater asset than knowledge learned once and done.
4. Tend to the Many Layers of Works Meaning Together
As we saw earlier, work gives not only income but meaning, structure, belonging, and identity together. So when preparing for change, it is good to look after all these dimensions together. When work is shaken, the gap in meaning and belonging can be as large as the gap in income. Cultivating in advance activities, relationships, and communities where you can feel meaning outside the workplace becomes another solid foundation for withstanding the shock of change.
5. Do Not Bear the Change Alone
No individuals effort alone can solve everything. Reskilling systems, the social safety net, and fair policy are a share that society must build together. Pushing the entire burden onto the individual with the line that adaptation is the individuals responsibility is not fair. Individual adaptation and social institutions are each insufficient on their own; only when the two mesh together can the shock of change be made bearable.
Does Reskilling Really Work
Almost every policy discussion about automation ultimately converges on the word reskilling: teaching new skills to those who lost work and moving them into new jobs. It sounds like a clean solution. But in reality, reskilling is a trickier task than it sounds. Looking honestly at this difficulty helps avoid both easy optimism and easy pessimism.
First, there is the problem of what to teach. In an age when technology changes fast, a skill that looks promising today may itself become a target of automation a few years later. So the view that it is safer to cultivate foundational capabilities and the ability to learn, which adapt broadly to change, rather than teaching one specific skill, gains force.
Second, there is the problem of who can learn. For a middle-aged worker who has done one kind of work all their life, supporting a family while learning the skills of a completely different field from scratch is by no means easy. The time, the cost, and the psychological burden are all considerable. The fact that a reskilling program exists in form does not mean it actually reaches everyone.
Third, there is the problem of whether jobs are really waiting afterward. If you painstakingly learn a new skill but there are no jobs in your area that use it, reskilling easily comes to nothing. This connects directly to the dilemma of the automated town we saw earlier.
This does not mean reskilling is meaningless. Historically, the expansion of education played a decisive role in absorbing the shock of the Industrial Revolution. As compulsory education took hold and public education expanded over the nineteenth and twentieth centuries, the rural workforce could move into the work of factories and offices. The key is that reskilling is neither a cure-all nor a useless thing. It is a tool that exerts its power only when it is well designed, sufficiently supported, and meshed with the demand for jobs, a tool that works only when several conditions are in place together.
Closing: What Should We Hold Instead of a Hammer
Let us return once more to that night in Nottinghamshire in 1811. The people with hammers wanted to stop their world from collapsing. Their fear was real, and the suffering they endured was real. We are in no position to simply sneer at them as foolish machine-breakers.
Yet at the same time, history shows that hammers could not stop the flow of technology. The Industrial Revolution eventually brought immense abundance, but it was not technology that made that abundance spread to everyone fairly and quickly enough; it was the choices of people. Labor law, public education, and social security systems did not fall from the sky; they were the product of a society that experienced the shock of change and, after long conflict and debate, made them.
AI will likely be the same. Whether AI becomes a threat or a tool is not predetermined in the technology itself. It depends on how we design that technology, how we use it, and how we divide its benefits and burdens.
So the question we ask must not stop at the fear-tinged question of whether AI will take my job. We should ask instead: How will we use this powerful tool to build a world where more people live doing better work? What we should hold instead of a hammer may, perhaps, be this better question.
In this essay we traveled a long way, starting from the Luddites hammer and passing through the copyist monks and the printing press, the spreadsheet and the accounting analyst, the ATM and the bank teller, and the telephone operator, to todays AI. If this long journey makes one thing clear, it is that technology does not decide the ending alone. The same technology resulted in abundance for the many in some societies and in inequality for the few in others. What made the difference was always the choices of people. In the age of AI, this point will not change. In the end, the answer lies not in the machine but in our hands.
Things to Think About Together
1. In your own occupation, what part will AI replace and what part will it augment? If you wrote the two down separately, what picture would emerge?
2. If universal basic income were introduced, how would the choices of your life change? And where do you think its funding should come from?
3. If the pattern that a particular generation bears the short-term pain while the next generation enjoys the long-term benefit is true, what should we do for the generation standing in the middle of the change?
4. If a person a hundred years from now looked back on us today, would they see us as Luddites with hammers, or as a generation that tamed its tools?
5. If the place you live were the automated town in this essay, which of the five options would you try first? And what would be the biggest obstacle to that choice?
6. Even assuming income is sufficiently guaranteed, where could you instead fill the meaning, structure, belonging, and identity you used to gain through work?
References
- Encyclopaedia Britannica, "Luddite" — https://www.britannica.com/event/Luddite
- History.com, "Who Were the Luddites?" — https://www.history.com/news/who-were-the-luddites
- James Bessen, "Learning by Doing: The Real Connection between Innovation, Wages, and Wealth" (Yale University Press) — https://yalebooks.yale.edu/book/9780300195668/learning-by-doing/
- Encyclopaedia Britannica, "Industrial Revolution" — https://www.britannica.com/event/Industrial-Revolution
- Stanford Encyclopedia of Philosophy, "The Ethics of Artificial Intelligence" — https://plato.stanford.edu/entries/ethics-ai/
- Encyclopaedia Britannica, "Universal Basic Income" — https://www.britannica.com/topic/universal-basic-income
- OECD, "The Future of Work" — https://www.oecd.org/employment/future-of-work/
- Encyclopaedia Britannica, "Printing" — https://www.britannica.com/technology/printing-publishing
- Encyclopaedia Britannica, "Johannes Gutenberg" — https://www.britannica.com/biography/Johannes-Gutenberg
- Encyclopaedia Britannica, "VisiCalc" — https://www.britannica.com/technology/VisiCalc
- Encyclopaedia Britannica, "Spreadsheet" — https://www.britannica.com/technology/spreadsheet
- World Economic Forum, "Future of Jobs Report" — https://www.weforum.org/reports/the-future-of-jobs-report-2023/
- International Labour Organization (ILO), "World Employment and Social Outlook" — https://www.ilo.org/digitalguides/en-gb/story/weso-trends
- McKinsey Global Institute, "Jobs lost, jobs gained" — https://www.mckinsey.com/featured-insights/future-of-work/jobs-lost-jobs-gained-what-the-future-of-work-will-mean-for-jobs-skills-and-wages
- Encyclopaedia Britannica, "Telephone operator" — https://www.britannica.com/technology/telephone
- Encyclopaedia Britannica, "Automated teller machine (ATM)" — https://www.britannica.com/technology/automated-teller-machine
현재 단락 (1/184)
On a winter night in 1811, shadows descended on a textile mill in Nottinghamshire, England. They car...