NextFin News - In a comprehensive 20,000-word essay that has sent shockwaves through Silicon Valley and global labor markets, Dario Amodei, CEO of the artificial intelligence safety startup Anthropic, has issued a stark warning regarding the trajectory of machine intelligence. According to IndexBox, Amodei claims that AI systems currently under development are on a path to surpass the cognitive capabilities of most humans across a vast array of professional tasks, potentially reaching this milestone as early as 2027. This evolution marks a transition from AI as a specialized tool to what Amodei describes as a "general labor substitute," capable of drafting complex legal memos, conducting deep data analysis, and managing intricate customer relations with greater efficiency than human counterparts.
The timing of this claim, released in late January 2026 and gaining significant traction by mid-February, coincides with a period of heightened economic anxiety. U.S. President Trump, inaugurated just over a year ago, faces a domestic landscape where AI-driven layoffs are no longer theoretical. Data from the first eleven months of 2025 revealed 55,000 AI-related job losses in the United States alone, representing 75% of all AI-related layoffs since 2023. Amodei’s forecast suggests that this is merely the beginning of a "snowball effect" that could see 50% of entry-level white-collar jobs vanish within the next five years. The warning is not limited to the United States; at the World Economic Forum in Davos, global leaders were urged to confront the reality that AI may do to office work what globalization did to manufacturing in the 1990s.
The core of Amodei’s argument rests on the concept of the "ability ladder." Historically, technological disruption has targeted routine manual labor, allowing workers to move "up" the ladder into cognitive, office-based roles. However, Amodei posits that AI is now climbing that ladder from the bottom to the top. When a customer service representative is displaced by a chatbot and attempts to retrain as a paralegal or a junior analyst, they find that AI has already occupied those roles. This creates a "closed escape route" for the workforce, potentially leading to the emergence of a permanent underclass of unemployed or very-low-wage workers whose cognitive output is no longer economically competitive.
From an analytical perspective, the impact of AI surpassing human cognitive capabilities is bifurcated by "adaptive capacity." According to a report from SF Standard, approximately 37.1 million U.S. workers are currently in high AI-exposure occupations. While 26.5 million of these individuals possess above-median adaptive capacity—meaning they have the savings, age, and transferable skills to pivot—roughly 6.1 million workers are at extreme risk. These are primarily office clerks, secretaries, and administrative assistants who lack the financial buffer or the specialized training to compete in an economy where "human-level" cognition is a commodity available for the price of a server subscription.
The economic implications are already manifesting in corporate strategy. By the end of 2026, 37% of business leaders anticipate replacing workers with AI, and 20% of large firms are specifically targeting middle management for automation. This trend suggests a hollowing out of the corporate hierarchy. While AI skills currently command a significant wage premium, acting as an equalizer for those who can master the technology, the long-term trend points toward a concentration of wealth. Amodei warns that if the gains from AI-driven productivity are not shared through concrete policy mechanisms, the resulting inequality could "break society."
Looking forward, the next 24 months will be a critical period for both the private sector and the U.S. government. U.S. President Trump’s administration is under increasing pressure to move beyond traditional job-creation rhetoric and address the structural reality of cognitive obsolescence. Potential future trends include the mandatory implementation of AI-displacement insurance for corporations and a renewed debate over Universal Basic Income (UBI) as a floor for the "unemployed underclass" Amodei describes. As AI moves into its "dangerous adolescence," the challenge will not be the technology’s failure to perform, but rather its overwhelming success in replicating the very cognitive traits that once defined human economic value.
Explore more exclusive insights at nextfin.ai.
