“Suddenly and suddenly”: Does the AI’s job displacement follow this pattern?

Whether it is by automating tasks, using as a co-pilot, or generating text, images, videos, and software from plain English, AI is rapidly changing the way we work. However, despite all discussions about the revolutionary work of AI, widespread labor displacement has not yet occurred.

It seems that this may be the calm before the storm comes. According to the recent World Economic Forum (WEF) Poll40% of employers expect their workforce to be between 2025 and 2030 anywhere AI can automate tasks. This statistics coincides with earlier predictions. For example, Goldman Sachs Research Report Two years ago, “the generated AI could make 300 million full-time jobs equivalent to automation, resulting in “severe damage” to the labor market.

according to For the International Monetary Foundation (IMF), “almost 40% of the world’s employment people are exposed to AI.” Brookings said last fall Report “More than 30% of workers can see at least 50% of career tasks being destroyed by AI agents.” A few years ago, Kai-Fu Lee, one of the world’s most important AI experts, said in 60 minutes, a few years ago, Kai-Fu Lee, one of the world’s most important AI experts. interview AI can replace 40% of global jobs in 15 years.

If AI is a destructive force, why don’t we see large layoffs?

Some question these predictions, especially since the work displacement of AI so far seems negligible. For example, October 2024 Challenger Report The layoffs that followed Jobs said there were fewer than 17,000 jobs in the U.S. due to AI in the 17 months from May 2023 to September 2024.

On the surface, this contradicts the horrible warning. But? Or does it mean that we are still in a gradual stage before a sudden transfer? History shows that technology-driven changes do not always occur in a stable linear way. Instead, it builds over time until a sudden transformation reshapes the landscape.

In the nearest one Hidden brain podcast About the turning point, researchers Rita McGrath Columbia University’s book Ernest Hemingway’s novel in 1926 The sun rises, too. When a character is asked how to go bankrupt, they reply, “Two ways. Gradually, then suddenly.” This could be a fable for the impact of AI on work.

This pattern of change—at first slow, almost undetectable, then suddenly undeniable—has been experienced in businesses, technology and society. Malcolm Gladwell calls it “Critical pointor the trend reaches critical mass moment, then greatly accelerates.

In cybernetics’ research (a study of complex natural and social systems), turning points occur when recent technology becomes so common that it fundamentally changes people’s lifestyles. In this case, change becomes self-reinforcement. This often happens when innovation and economic incentives are aligned, making change inevitable.

Gradually, then suddenly

Although the employment impact of AI is a new birth (so far), this is not the case with AI adoption. In the new Poll Written by McKinsey, 78% of respondents said their organizations use AI in at least one business function, up more than 40% from 2023. Other research findings 74% of corporate C-Suite executives Now, you have more confidence in AI business advice than colleagues or friends. The study also shows that 38% trust AI to make business decisions for them, while 44% succumb to AI inferences about their own insights.

It’s not just business executives that have increased their use of AI tools. New charts from investment firm Evercore depict the increase in usage for all ages over the past nine months, regardless of its application.

This data reveals the widespread adoption and growing adoption of AI tools. However, according to another McKinsey, true enterprise AI integration is still in its infancy – only 1% of executives describe their Gen Gen Rollouts as mature. Poll. This suggests that while AI adoption is surging, companies have not fully incorporated into their core operations in ways that may replace work with the scale of work. But that may change soon. If economic pressures intensify, companies may not gradually adopt the luxury of AI and may feel the need to automate quickly.

Canary in the coal mine

One of the first working categories that could be hit by AI is software development. There are a large number of AI tools based on large language models (LLM) to enhance programming, and soon the feature can be fully automated. Dario Amodei, CEO of Humanitarian explain Recently said on Reddit: “We are the world where AI writes 90% of code for 3 to 6 months.transparent

This trend is becoming clearer, as evidenced by startups in the winter incubator queue in 2025 Y combination. Managing partner Jared Friedman said 25% of the startup 95% of the code base generated by AI. He added: “A year ago,[the company]could have built its products from scratch, but now 95% are built from AI.”

The basic code generation of LLM, such as Claude, Gemini, Grok, Llama and Chatgpt, is developing rapidly and performing well in a series of quantitative benchmarks. For example, OpenAI’s inference model O3 missed only one question from the 2024 US Invitational Mathematics Exam, with a score of 97.7%, and won 87.7% of the awards on GPQA Diamond, which has graduate-level biology, physics, and chemistry problems.

What’s even more amazing is the qualitative impression of the new GPT 4.5, such as confluence postal. GPT 4.5 correctly answers broad and vague tips that other models cannot. This doesn’t seem outstanding, but the author notes: “This trivial communication is the first conversation with LLM, and we walked away and thought,’ That It feels like general intelligence.

Critical point

While software engineering may be the first knowledge worker profession in the face of widespread AI automation, this is not the last one. Many other white-collar work covering research, customer service, and financial analysis may also be exposed to AI-driven disruption.

What could lead to a sudden shift in AI adoption in the workplace? History shows that recessions often accelerate technology adoption, and the next recession may be the gradual shift in AI’s impact on jobs to a sudden turning point.

During the economic downturn, businesses face pressure to reduce costs and increase efficiency, making automation more attractive. The labor force becomes more expensive compared to technology investments, especially when companies require less human resources to do more. This phenomenon is sometimes called “forced productivity.” For example, the Great Recession between 2007 and 2009 has made significant progress in automation, cloud computing and digital platforms.

If the recession is achieved in 2025 or 2026, companies facing pressure to reduce employees may turn to AI technologies, especially LLM-based tools and processes, as a strategy to support efficiency and productivity. This may be more obvious and more sudden considering the business’ concerns about the lagging of AI adoption.

Will there be a recession in 2025?

It is always difficult to tell when a recession occurs. Chief Economist at JP Morgan Estimated 40% chance. Former Treasury Secretary Larry Summers said it could be about 50%. The betting market is aligned with these views, with a more than 40% chance of a recession predicting that it will occur in 2025.

If the recession occurs in late 2025, it can indeed be described as an “AI recession.” However, AI itself will not be the reason. Instead, economic necessity may force companies to accelerate automated decision-making. This is not a technical necessity, but a strategic response to financial pressure.

The extent of AI impact will depend on several factors, including the speed of technology maturity and integration, the effectiveness of workforce retraining programs, and the adaptability of businesses and employees to an evolving landscape.

Whenever this happens, the next recession may not only lead to temporary unemployment. Companies that have been trying out AI or adopting it in limited deployments may suddenly find that automation is not optional, but survival is crucial. If this happens, it could mean a permanent shift towards a more driven workforce.

As Salesforce CEO Marc Benioff said Recent income Please call: “We are the last generation of CEOs who only manage humanity. Every CEO will manage humanity and the agency together. I know that’s what I’m doing. … You can see it in the global economy as well. I think that without more human labor, productivity will increase, which is good because human labor will not increase in global labor.”

Many of the biggest technological changes in history coincide with the economic downturn. AI may be the next one. The only question left is: Will AI not only increase jobs in 2025, but also start replacing them?

Gradually, then suddenly.

Gary Grossman is the EVP of Technology Practice Edelman and global leadership of Edelman AI Center of Excellence.

Daily insights on VB daily business use cases

If you want to impress your boss, VB Daily can serve you. We provide you with insights about the company’s work in developing AI, from regulatory to actual deployment, so you can share your insights on the maximum ROI.

Read our Privacy Policy

Thanks for your subscription. See more VB newsletter here.

An error occurred.