AI and jobs: One leader warns it may already be too late

Dario Amodei of Anthropic warns that AI is advancing rapidly and could be dominant within two to three years, and he says AI is already writing much of the code at his company.

Dario Amodei, co-founder of Anthropic, published an essay warning that artificial intelligence is advancing quickly and could perform most tasks in the future. He argues enterprise AI adoption is accelerating and may produce major market shifts within a short time frame. Amodei also reports that much of the code at Anthropic is now being written by AI, creating a feedback loop in development. The essay appeared alongside revisions to the Claude Constitution announced at the World Economic Forum in Davos. Key developments: - Amodei warns AI could eventually perform a very wide range of tasks and says society needs to prepare for broad economic effects. - He states enterprise AI adoption is growing faster than prior technologies and could lead to significant market disruption within two to three years. - Anthropic reports AI is writing much of the company’s code, which the company says is accelerating its progress and creating a self-reinforcing development loop. - Anthropic announced revisions to the Claude Constitution at Davos, described as efforts to align AI behavior with human values amid growing regulatory interest. Summary: Amodei frames rapid AI capability gains as a potential source of major economic and career disruption, and he notes both startup activity and new business models that could follow. He points to internal changes at Anthropic as an example of accelerating automation. The broader policy and labor outcomes he discusses are not resolved. Undetermined at this time.