Tech Desk: In a move that has sparked significant debate, AI powerhouse Anthropic has confirmed the operation of a wet biology lab in the Bay Area, where it leverages its AI models to conduct physical experiments. This revelation comes as leading AI figures, including Anthropic CEO Dario Amodei, continue to issue dire warnings about AI’s potential for catastrophic risks, including bioterrorism and even human extinction.

The company states its “Anthropic wet biology lab” focuses on fundamental biology, not drug discovery, aiming to avoid direct competition with pharma partners. This strategic emphasis follows its acquisition of stealth AI biotech Coefficient Bio in April and recent initiatives like the Life Sciences Verification Program, designed to give vetted researchers access to its powerful models for accelerating protein design and biomolecular modeling.

However, the establishment of a physical lab dedicated to biological experiments stands in stark contrast to the company’s own public warnings. Former Anthropic researcher Jacob Coxon warned that builders earnestly believe AI could “kill us all,” a sentiment echoed by the company’s alignment lead giving a greater than 10% chance of humanity’s extermination within a decade.

Investor Chamath Palihapitiya wryly noted the juxtaposition on X, questioning the wisdom of a group warning of global demise simultaneously building such a lab. This dichotomy highlights the complex and often contradictory landscape of advanced AI development, where the pursuit of groundbreaking solutions coexists with profound ethical and safety dilemmas.