Research map / Large language models / AI and society

Position: AI Safety should prioritize the Future of Work

ICML 2025 · Large language models › AI and society · approach: Human-centered trust and agency

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Position paper arguing AI safety should prioritize the future of work, outlining economic risks of generative AI to labor and recommending pro-worker governance, copyright licensing, and watermarking.

Problem

Current AI safety research focuses on harmful content, manipulation, and catastrophic risks, overlooking the medium- and long-term effects of generative AI on labor markets, shared prosperity, learning, and creative work.

Approach

Reframes the future of work as a core AI safety concern, linking intertemporal consumption theory, rent-seeking, extractive institutions, and Coasean transaction costs to AI-driven labor displacement, and proposes collective licensing, royalty-based compensation, watermarking, and worker-inclusive governance.

Key results

• Cites Demirci et al. (2024) finding a 21% decrease in writing and coding job posts after ChatGPT's introduction and a 17% decrease in image creation jobs after image-generating AI.
• Cites Peng et al. (2023) finding programmers were 55.8% faster when assisted by an AI programmer.
• Cites Eloundou et al. (2023) estimate that around 80% of the U.S. workforce could have at least 10% of work tasks affected by LLMs.

Surprising findings

• Argues that AI detectors become ineffective once models are fine-tuned, and that watermarking adoption suffers from a prisoner's dilemma among firms.

Limitations

• No new empirical experiments; arguments rest on cited literature and economic theory.
• Recommendations are high-level policy proposals without concrete implementation evaluation.

Experimental setup

Builds on: Life-Cycle Hypothesis Permanent Income Hypothesis Shapley Royalty Share

Closest papers by meaning

• Position: Humanity Faces Existential Risk from Gradual Disempowerment (ICML 2025 · similarity 0.86)
• Regulating AI: Applying Insights from Behavioural Economics and Psychology to the Application of Article 5 of the EU AI Act (AAAI 2024 · similarity 0.84)
• Position: If Innovation in AI systematically Violates Fundamental Rights, Is It Innovation at All? (NeurIPS 2025 · similarity 0.83)
• Addressing Digital and AI Skills Gaps in European Living Areas: A Comparative Analysis of Small and Large Communities (AAAI 2024 · similarity 0.83)
• JobBench: Aligning Agent Work With Human Will (NeurIPS 2026 · similarity 0.82)
• A Mathematical Framework for AI-Human Integration in Work (ICML 2025 · similarity 0.82)