Chairman of Coursera: AI companies inciting employment panic just to sell products, software industry still actively hiring

Scholar Andrew Ng refutes the narrative that artificial intelligence will lead to a wave of unemployment. He points out that software recruitment remains strong, and much of the panic rhetoric is just corporate interest packaging, with more job opportunities expected in the future.

On May 12, Coursera Chairman, well-known AI scholar, and founder of DeepLearning.AI Andrew Ng (吳恩達) posted on X and The Batch newsletter asserting that “AI will not trigger a jobpocalypse,” directly countering the mainstream AI unemployment panic narrative. According to Ng’s original tweet, this post received over 2,600 likes and was one of the most talked-about opinion pieces in the AI field that week.

Ng’s core argument: software engineering recruitment remains strong, unemployment rate stays at 4.3%

Ng uses three sets of concrete data to counter the narrative that “AI will cause massive unemployment”:

• Software engineering is the industry most affected by AI tools (rapid progress in coding agents), yet recruitment for software engineers remains vigorous
• Despite rapid AI advancements, the U.S. current unemployment rate stays at a healthy 4.3%
• From historical experience: the number of new jobs created by AI significantly exceeds the number of jobs it replaces, consistent with past technological waves

Ng straightforwardly states: “AI—like any other technology—does impact jobs, but exaggerating stories of large-scale unemployment is irresponsible and harmful. We should put an end to such narratives.”

Why is the “AI unemployment” narrative so popular? Ng points out three incentives

Ng highlights three structural incentives explaining why this narrative continues to be amplified:

First, leading AI labs have strong incentives to promote the story that “AI can replace workers”—if a technology can replace many employees, it appears more valuable. In extreme cases, labs even push sci-fi scenarios like “AI takeover leading to human extinction.”

Second, SaaS software companies typically charge $100–$1,000 per user annually, but if AI can replace a $100k annual salary worker or boost employee productivity by 50%, they can charge $10k, which still seems reasonable. Pricing anchored on “employee salary” rather than “typical SaaS prices” allows AI companies to charge more.

Third, companies have strong incentives to frame layoffs as “due to AI adoption”—telling stories that AI enables them to achieve higher productivity with fewer employees is more respectable than admitting they overhired during low-interest, government-stimulated periods like during the pandemic.

Historical comparisons: nuclear energy, overpopulation fears, low-fat diets

Ng cites three historical cases to illustrate how societal narratives can persist for years but are actually disconnected from reality:

• Safety fears about nuclear power, leading to long-term underinvestment in nuclear energy
• 1960s “population bomb” panic, resulting in strict population control policies in many countries
• Concerns about dietary fats, leading governments to promote high-sugar diets for decades

Ng states: “Now mainstream media are beginning to question the jobpocalypse openly. I hope the influence of such stories will gradually fade, just like fears of AI causing human extinction.”

Ng’s reverse prediction: AI jobapalooza

Ng offers a counter-prediction to the “AI wave of unemployment”—“AI jobapalooza” (AI grand festival):

• A large number of good AI engineering jobs will emerge, and the overall job market outlook remains optimistic
• AI engineer roles will differ from traditional software engineering, with new jobs scattered across companies outside the “big traditional developers”
• Skills required for non-AI roles will change due to AI, necessitating more people to develop “AI proficiency”

Tech news observation: Ng’s timing coincides with this week’s launch of OpenAI’s Deployment Company, the joint venture between Anthropic and Blackstone, JPMorgan and BlackRock’s tokenized funds, and other AI commercialization accelerations. Ng does not deny AI is reshaping work but opposes exaggerated predictions of “rapid large-scale unemployment.” For Taiwanese readers, Ng’s arguments can be used to evaluate which mainstream panics might be amplified by vested interests and which pose real risks.

Follow-up events to watch include: whether mainstream media truly shifts toward questioning the jobpocalypse, responses from leaders of labs like Anthropic and OpenAI to Ng’s criticism, and whether labor market data in late 2026 (especially tech sector recruitment numbers) support Ng’s forecast.

• This article is reprinted with permission from: 《鏈新聞》
• Original title: 《Andrew Ng：「AI 不會引發失業大潮」、軟體業徵才仍熱》
• Original author: Elponcrab