Anthropic CEO: AI to Displace Jobs, Criticizes Competitors

Anthropic CEO Dario Amodei has warned that artificial intelligence could replace up to 50% of entry-level white-collar jobs as AI systems achieve "Ph.D.-level" intelligence. Speaking at a recent summit in New York, Amodei also critiqued the aggressive investment strategies of some AI competitors, implicitly targeting OpenAI and its CEO, Sam Altman.

BlackRock CEO Larry Fink, also present at the event, echoed concerns about the AI industry, stating that the race for AI dominance would produce "super winners" and "super losers."

AI's Impact on Employment and Societal Values

Amodei outlined a "corporate-government-society" three-tier model to address the impact of AI on employment. While acknowledging that retraining programs are not a complete solution, he stressed their necessity and called for proactive government intervention. He suggested that AI might eventually transform societal values, shifting the focus away from work as life's central purpose and enabling a social structure where "economic survival" is not the sole objective.

He differentiated AI from previous technological revolutions like the internet or 5G, stating that the market alone would not regulate its effects. Amodei emphasized that those who truly understand AI are concerned about issues beyond national security, including model alignment and employment impacts. He likened dismissing AI safety concerns to "driving on the road, but first I'll remove the steering wheel." In May, Amodei had previously warned that AI could lead to unemployment rates of 10-20% within five years due to the disappearance of entry-level jobs, clarifying that his warnings aimed to prompt solutions rather than spread doomsday prophecies.

Industry Outlook and Investment Trends

The New York Times' annual DealBook summit featured discussions on the direction of the AI industry and the potential for an "AI bubble." Despite hundreds of billions of dollars invested by tech giants like Meta, Alphabet, and Microsoft into AI infrastructure this year, and 49 U.S. AI startups raising at least $100 million, attendees acknowledged that the rapid growth could lead to instability.

Fink told DealBook founder Andrew Ross Sorkin that the industry would see "epic 'headline blow-ups'" alongside its successes. However, Fink also noted strong demand, reporting that "hyperscalers" such as Amazon, Google, and Microsoft privately complain about a lack of computing power.

Amodei's Critique of Competitors

Amodei described the AI industry as inherently high-risk due to the substantial costs of building data centers. He warned that even minor misjudgments in timing or strategy could have disastrous consequences for companies, even if the technology delivers on its promises.

He stated that Anthropic's strategy prioritizes "stability first," focusing on partnerships with large enterprise clients and conservative computing power investments. Amodei hinted that some competitors are pursuing an "all in" approach, which he characterized as reckless gambling. While not naming specific companies, his remarks were interpreted as a critique of OpenAI and Altman, his former employer and boss. He specifically referenced a leaked OpenAI memo about an internal "red code" alarm in response to a new Google model, stating, "We never do anything like 'red code'."

Anthropic's Differentiated Strategy

Amodei explained that Anthropic's calm approach stems from its differentiated strategy, focusing on enterprise users (B2B) rather than competing in the consumer market (B2C). He noted that Anthropic optimizes its models for business needs, prioritizing code capabilities, high-intensity intellectual activity, and research.

Anthropic's Claude Opus 4.5 model, released last month, is designed for generating computer code and processing office documents. The company plans to expand into finance, biomedicine, retail, and energy sectors. Amodei expressed skepticism about the frantic spending by larger tech giants to secure top AI positions, questioning whether the economic value growth can keep pace with the expenditure. He stated that Anthropic aims to "manage risk as responsibly as possible."

Path to Artificial General Intelligence

Amodei believes that scaling is likely the path to Artificial General Intelligence (AGI). He observed that minor adjustments consistently strengthen AI models, describing this as the core driver of progress. He suggested that AGI would not appear as a sudden "magic moment" but rather as a linear progression of increasing model capabilities.

He noted that Claude can now write most code, with engineers primarily modifying its output. Amodei predicted that future models would continue to be smarter, leading to a doubling of economic value. According to recent reports, Anthropic is preparing for an IPO as early as 2026 and is currently negotiating a new funding round at a valuation of $300-350 billion.