AGI countdown has begun: Anthropic CEO predicts large-scale disappearance of white-collar jobs

Anthropic CEO Dario Amodei made a major prediction at the Davos Forum: Artificial General Intelligence (AGI) with superhuman capabilities may emerge between 2026 and 2027. This not only signifies that AGI has moved from theoretical discussion to a real countdown but also, more importantly, he explicitly stated that entry-level white-collar jobs will disappear on a large scale. Meanwhile, Anthropic’s funding scale and revenue growth are also validating industry confidence — annualized revenue has exceeded $9 billion, and new funding rounds have surpassed $20 billion. The company is betting real money on the imminent arrival of AGI.

AGI Prediction: A Contrast of Perspectives from Two AI Leaders

Amodei’s prediction is quite aggressive. He pointed out that AI development has entered a self-accelerating phase — a key statement indicating that the cycle of AI improving AI is forming. Within Anthropic, most programmers are responsible for editing model-generated code rather than coding from scratch. This reflects a phenomenon: automation in software development is no longer a future concept but an ongoing reality.

In contrast, Google DeepMind CEO Demis Hassabis’s prediction is more conservative. He believes scientific creativity remains difficult for AI to replicate, estimating only a 50% chance of achieving AGI before 2030. The difference between these two views is quite interesting — one focuses on the acceleration of engineering automation, the other on the difficulty of breakthroughs in creativity and science.

Structural Impact on the Labor Market

Both CEOs warn that governments are underestimating AI’s impact on the labor market. Specifically:

White-collar jobs are most at risk

Entry-level white-collar work is easiest to automate — these roles often involve standardized processes, predictable tasks, and repetitive work. From programmers to analysts to copywriters, these positions face the risk of being replaced by AI. Amodei’s statement that these jobs will “disappear on a large scale” is not an exaggeration but an intuitive description of automation trends.

The time window is very tight

If AGI truly emerges between 2026 and 2027, then governments need to formulate policies, companies must adjust organizational structures, and the workforce needs to upgrade skills — all within less than two years. This explains why both CEOs emphasize that “the current system’s preparation time is urgent.”

Anthropic’s Market Position and Funding Advantages

From a business perspective, Anthropic’s data is more convincing:

Revenue growth exceeds expectations

According to the latest reports, Anthropic’s annualized revenue has grown from $4 billion in July 2025 to over $9 billion by the end of 2025, more than doubling — this is not just funding figures but actual income, indicating that Claude’s commercialization is far beyond expectations.

Record-breaking funding scale

The latest funding round exceeded $20 billion, with investors including Coatue Management, Singapore’s GIC, Iconiq Capital, as well as previous commitments from Nvidia and Microsoft totaling $15 billion. This scale of funding reflects investors’ high confidence in Anthropic’s competitiveness in AGI.

Ecosystem expansion accelerates

Humans&, founded by former Anthropic, xAI, and Google employees, completed a $480 million seed round with a valuation of $4.48 billion. This shows that Anthropic’s technological accumulation and talent pool are sufficient to incubate new AI startups, forming an ecosystem.

Institutionalization of Safety and Ethics

It is noteworthy that Anthropic has also updated Claude’s constitutional framework — an 80-page ethical guideline that adds more details on safety and ethics. This move coincides with the announcement of the AGI timetable, implying that while pursuing capability enhancement, Anthropic is also strengthening safety foundations. This balance may be one of the sources of investor confidence.

Summary

AGI is no longer a matter for 20 years later but a reality that could happen within 2-3 years. While Amodei’s prediction is aggressive, the surge in funding, revenue growth, and technological progress suggest it is not unfounded. The real challenge lies not in technology but in society — governments, enterprises, and the workforce are not prepared for such a level of change. In the next two years, we may see intense discussions on AI governance, labor policies, and social welfare systems. For individuals, this is also a signal: if your work mainly involves routine tasks, now might be the time to reconsider your career direction.