According to @lisaabramowicz1, Anthropic’s chief scientist Jared Kaplan said AI systems could handle most white-collar work within two to three years, as reported by The Guardian on December 2, 2025. Source: @lisaabramowicz1 on X; The Guardian. The Guardian interview also highlights Kaplan’s view that future AI models may increasingly train themselves, indicating a faster capability ramp defined by the lab’s own research trajectory. Source: The Guardian. For traders, this explicit 2–3 year horizon offers a concrete benchmark to reassess positioning across AI-exposed equities and AI narratives in digital assets, grounded in a senior lab scientist’s guidance rather than marketing claims. Source: @lisaabramowicz1 on X; The Guardian.