Will AI Take Jobs or Transform Them? The "Adaptability" Competition in the Era of 800 Million People

Is the destruction of jobs by AI really imminent, or are we merely in the process of transitioning to new forms of work, as we have with past technological innovations?

A new analysis from Bank of America, reported by Investing.com, suggests that AI will not replace jobs worldwide all at once but will significantly transform the nature of work. According to the report, approximately one in four jobs globally, affecting around 840 million people, could be impacted by AI. However, the report emphasizes that the future is not as simple as "entire professions disappearing." Instead, certain tasks within jobs will be automated, leading to a reallocation of human roles.

This perspective can bring a degree of calm to the heated debate surrounding AI. However, it is not merely reassuring. Even if AI does not completely eliminate jobs, it will significantly impact wages, promotions, hiring, education, and the distribution of corporate profits. The issue is shifting from "whether jobs exist" to "who benefits from AI and who gets left behind."

It's Not "Jobs" That Change, But "Tasks"

The most important aspect of this report is its focus on the impact of AI at the task level rather than the job level.

For example, consider the job of an accountant. AI might automate the classification of invoices, number matching, and the creation of standard reports. However, it cannot fully handle exception processing, client coordination, necessary explanations for management decisions, or alignment with legal and tax matters. Thus, the profession of accounting does not disappear instantly; rather, the proportion of tasks performed daily by accountants changes.

This applies to many white-collar jobs such as sales, marketing, human resources, customer support, legal, healthcare, education, software development, and media production. Tasks like writing, summarizing, translating, organizing data, researching, code assistance, image generation, and responding to inquiries are areas where generative AI excels. However, understanding the context on the ground, making final judgments, taking responsibility, interpersonal communication, and organizational coordination remain more likely to be handled by humans.

Given this structure, employment anxiety in the AI era cannot be measured by whether one's job title is taken over by AI. More important is determining "which tasks in your job can be shortened by AI and which tasks remain as human value."

AI Will Create New Industries, But the Transition Period Will Be Painful

The Bank of America report compares the spread of AI to the Industrial Revolution and the proliferation of computers. Indeed, looking back at history, new technologies have often destroyed existing jobs while simultaneously creating new ones. The introduction of factory machinery reduced some manual labor but expanded new employment in manufacturing management, machine maintenance, logistics, sales, and quality control. Similarly, PCs and the internet reduced typist and paper-based clerical work while creating roles such as IT administrators, web designers, digital advertising, e-commerce operations, and data analysis.

AI holds the same potential. New job domains such as AI implementation consultants, prompt design, AI auditing, data quality management, verification of generated content, AI ethics, model operation, internal AI education, and AI-driven business design are already beginning to expand. People who can use AI not only speed up tasks but also enable organizational designs that achieve significant results with fewer people.

However, another historical fact cannot be overlooked: the benefits of technological innovation are not distributed fairly from the start. Companies with new technologies, capital, management, and highly skilled personnel benefit first, while workers in the transition period often face wage stagnation, redeployment, hiring freezes, and the burden of retraining.

In other words, while employment may be reorganized in the long term, in the short term, there will be groups that bear concentrated pain. What is required in AI-era policy and corporate management is not just optimism that "new jobs will eventually be created," but a clear view of who will bear the losses along the way.

The Real Risk of Widening Inequality

A particularly heavy point in this report is the potential for AI to widen inequality rather than cause mass unemployment. There is a concern that the fruits of productivity improvements brought by AI may skew towards companies and capital owners rather than workers.

From the corporate perspective, AI is attractive. It can accelerate document creation, customer service, analysis, development, and advertising production while keeping labor costs down. If the productivity of existing employees increases, the same revenue can be achieved with fewer people. If businesses can operate with reduced new hires, corporate profits improve.

However, from the workers' perspective, the situation becomes more complex. Even if productivity increases with AI, it does not necessarily translate into wage increases. Instead, judgments such as "since efficiency can be achieved with AI, we won't increase staff," "routine tasks previously assigned to juniors can be handled by AI," and "there's no room to train inexperienced people" may spread, potentially thinning entry-level employment.

This is serious for younger generations. In many workplaces, juniors are not entrusted with advanced judgment tasks from the start. They gain experience through relatively routine tasks like document creation, minute-taking, research, simple analysis, customer service, code corrections, and drafting. However, if these entry-level tasks are replaced by AI, the very steps for young people to learn practical skills are lost.

The phrase "AI assists humans" might be correct. However, those who can be assisted are likely those who already possess judgment and expertise. The problem lies in where those who are yet to acquire expertise will gain their experience.

Why Women, Young People, and Highly Educated Groups Are More Affected

AI's impact does not fall evenly on all workers. Related analysis by the International Labour Organization also shows that occupational exposure to generative AI varies by national income level, gender, and occupation. Particularly, occupations involving many digitalizable tasks, such as clerical, administrative, and document processing, are more susceptible to impact.

Therefore, the changes brought by AI differ from the traditional image of "factory workers being replaced by robots." Instead, they strongly affect some white-collar, clerical, professional, and creative jobs. Jobs involving handling text, organizing information, and making patterned judgments often overlap with the strengths of generative AI.

There is also a suggestion that clerical, support, and administrative roles, where many women are employed, are more susceptible to AI's impact. If AI implementation proceeds solely for corporate efficiency, it risks widening existing gender gaps. Conversely, if combined with retraining and flexible working arrangements, it could open new employment opportunities for those whose careers were interrupted by childcare or caregiving.

AI may seem like a neutral tool, but its implementation is not neutral. Decisions about which departments to invest in, who to train, and whose jobs to cut reflect corporate judgment and societal systems.

On Social Media, "Doubt" Overpowers "Reassurance"

Reactions on social media to reports like "AI will not completely take jobs" are not monolithic. Broadly speaking, optimism, skepticism, and anxiety intersect.

Optimistic reactions often include perspectives such as "AI changes daily tasks rather than eliminates professions," "it's ultimately about learning new tools," and "there was similar anxiety with PCs and the internet, but society adapted." Especially among technologists and those already using AI, there are voices that see AI not as a threat but as a productivity-enhancing partner. People who feel the practical effects in tasks like code assistance, drafting, time-saving in research, and brainstorming tend to be skeptical of the view that "everyone will lose their jobs."

On the other hand, there is strong distrust. On social media, some suggest that "it's not that AI is genuinely adding value and reducing people, but that budget allocations for AI investments and explanations to shareholders are driving personnel reductions." In other words, there is suspicion that the phrase "AI implementation" is being used as a pretext for business rationalization, beyond the capabilities of AI itself.

This reaction is important. Even if AI cannot yet fully replace humans, if companies declare "efficiency through AI" and reduce hiring or staff, the impact on the labor market becomes real. Whether it is technically possible and whether it is a management decision to reduce staff are separate issues.

Furthermore, voices expressing anxiety remain strong. Particularly among young people, job seekers, creative professionals, clerical workers, and customer support roles, there are concerns such as "will entry-level jobs disappear before gaining experience?" and "will only a few who can use AI benefit while the rest face lower wages?" This is not just emotional rhetoric. The potential pressure on entry-level jobs is also touched upon in the report.

What emerges from social media discussions is that people are not only afraid of AI itself. Rather, they are watching how companies manage labor costs, change hiring practices, and distribute profits using AI. Anxiety about AI is both a fear of technology and a distrust of management and distribution.

The Divide Between "Those Who Can Use AI" and "Those Used by AI"

What will become important in future workplaces is whether we can move from merely using AI as a convenient tool to redesigning the work itself.

Those who can use AI are not just people who can input prompts. They are people who can deconstruct their work, decide which tasks to delegate to AI, where humans should make judgments, and which outputs need verification. They are people who do not take AI's answers at face value, can spot errors and biases, and can adjust to meet objectives.

For example, for a writer, it is not enough to have AI write articles. Value lies in theme setting, understanding the audience, structuring, fact-checking, offering unique perspectives, and adjusting expressions. For an engineer, value lies not just in code generation but in requirements definition, design, security, maintainability, and team development decisions. For sales, value lies not just in automating email content but in reading the customer's true intentions and building relationships.

In the AI era, the weak approach is to consider tasks that AI can perform as one's own value. Conversely, the strong approach is to direct the time saved by AI towards higher-level judgment, creativity, interpersonal coordination, and strategic planning.

The Responsibility of Companies

It is dangerous to place the burden of employment changes due to AI solely on individual efforts. Companies also bear significant responsibility.

Firstly, the purpose of AI implementation needs to be clarified. Is it merely for staff reduction, improving employee productivity, or enhancing customer experience? If introduced with ambiguous objectives, it will confuse the field and breed distrust among workers.

Secondly, investment in retraining is essential. If AI is to be implemented, it is necessary to teach affected employees how to use it, verify methods, and think about business redesign. If AI training is limited to some managers and digital personnel, internal disparities will widen.

Thirdly, the mechanism for nurturing young people must be redesigned. If AI handles routine tasks, young people need to be given experience closer to judgment tasks from the start. This requires mentoring systems, simulation-based training, practical exercises using AI, and phased responsibility transfer.

Fourthly, how to distribute the fruits of productivity improvement is questioned. If profit margins rise with AI but wages do not grow and employment becomes unstable, societal backlash will intensify. AI implementation should not only enhance corporate value but also link to employee growth and rewards.

What Individuals Can Do Now

The actions individuals should take in the AI era go beyond merely "having expertise that AI cannot surpass." Rather, it is necessary to update expertise with AI as a premise.

First, break down your job into tasks. List daily tasks and divide them into information gathering, drafting, analysis, verification, judgment, interpersonal coordination, and decision-making. Then, determine which parts can be delegated to AI, which parts should be handled by humans, and where AI and humans can collaborate.

Next, develop the ability to verify AI outputs. AI can produce plausible errors. Therefore, foundational skills such as industry knowledge, legal regulations, numerical sense, detecting inconsistencies in writing, and customer understanding become more important. As AI becomes more widespread, those without foundational skills will not notice AI's errors, while those with foundational skills can use AI as an amplification tool.

Furthermore, shift your value from "amount of work" to "outcomes." The value of spending long hours creating documents itself decreases. Instead, what is questioned is which questions were raised, which judgments were supported, and which outcomes were achieved.

Finally, a continuous learning attitude is indispensable. The report also states that future employment stability is determined by "adaptability" rather than "job stability." While this is a harsh statement, it is also realistic. Instead of looking for unchanging jobs, the ability to reshape one's role in response to changes is necessary.

AI Is Not the End, But a Renegotiation of Labor

Discussions surrounding AI often fall into a binary opposition of "human or machine." However, what is actually happening is the redefinition of human work.

AI writes text. It writes code. It creates images. It summarizes meetings. It responds to customers. It organizes data. It takes over many tasks that humans have spent time on. Therefore, humans are more strongly questioned about "what is the purpose," "what is correct," "who is it for," and "which results are accountable."

AI may not take all jobs. However, it certainly changes where the value of work is placed. Tasks that were previously valued become rapidly cheaper, while judgment, editing, coordination, ethics, and understanding of the field, which were previously less visible, become important.

It is natural that voices of anxiety on social media do not disappear. People fear not only the technology of AI but also the sudden devaluation of their experience, the loss of opportunities for young people to grow, companies profiting alone, and society pushing the burden of change onto individuals.

Therefore, the real issue in the AI era is not "whether AI will take jobs" but "for whom the productivity created by AI will be used."

If companies use AI solely as a tool for staff reduction, societal distrust will spread. If workers continue to reject AI, they will be left behind by change. If governments and educational institutions delay retraining, inequality will become entrenched.

AI is not the end of employment. However, it is not a future where one can feel secure without doing anything. What begins now is a large-scale renegotiation of human work. What divides those who can participate in this negotiation is not fear or optimism but the ability to interpret change and reshape one's role.

Source URL

Investing.com: Referencing Bank of America's report, which views AI as "reshaping" rather than "replacing" jobs, the potential impact on approximately one in four people globally, around 840 million, and the implications for inequality and entry-level positions.
https://www.investing.com/news/technology-news/ai-set-to-reshape--not-replace--global-jobs-new-report-finds-4654906

Investing.com: Related article referencing Bank of America citing ILO data on the potential impact of generative AI on approximately 838 million jobs.
##HTML