Toward Human-Centered AI-Assisted Terminology Work

2025-12-21 · Source: cs.CL updates on arXiv.org · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Terminology Management · Depth: Advanced, quick

Summary

Antonio San Martin's paper, "Toward Human-Centered AI-Assisted Terminology Work" (arXiv:2512.18859v3, revised 17 Jun 2026), explores the transformative impact of Generative AI on terminology work. While AI offers automation opportunities, it also raises concerns due to its unreliability, including errors, hallucinations, and bias, making human terminologists essential for data accuracy. The paper advocates for a human-centered AI framework to maximize benefits and mitigate risks, asserting that high automation can coexist with meaningful human control. It examines this integration across three dimensions: the augmented terminologist, ethical AI considerations, and human-centered design principles. Specifically, it addresses how AI reshapes the terminologist's role, affects professional values, necessitates managing AI-generated bias, and requires designing tools around human needs. The author concludes that this human-centered orientation is vital to ensure AI strengthens terminology's role in specialized communication.

Key takeaway

For AI Ethicists and terminology managers evaluating generative AI integration, you should prioritize human-centered design to ensure AI augments, rather than replaces, expert terminologists. Focus on developing tools that manage AI-generated bias and preserve human agency, as large language models remain unreliable for critical terminological accuracy. Your strategy must balance automation efficiency with the indispensable need for human oversight to maintain data reliability and professional values.

Key insights

Human-centered AI integrates automation with human oversight to enhance terminology work while mitigating generative AI's risks.

Principles

• AI should augment, not replace, human expertise.
• High automation and human control are compatible.
• AI tools must be designed around human needs.

Method

The paper examines AI integration through three dimensions: the augmented terminologist, ethical AI, and human-centered design, focusing on role reshaping, professional values, bias management, and tool design.

In practice

• Manage AI-generated bias in terminological data.
• Design AI tools to preserve terminologist agency.
• Prioritize human well-being in AI deployment.

Topics

• Human-Centered AI
• Terminology Work
• Generative AI
• AI Ethics
• Large Language Models
• Automation

Best for: Research Scientist, Domain Expert, AI Ethicist, AI Scientist

Related on AIssential

• 🔮 Where the human ends and AI begins — AI blurs human agency, necessitating deliberate intent to preserve core human competencies amidst increasing automation and synthetic content.
• Societal Impacts — AI research must integrate human values, anticipate misuse, and inform policy to ensure beneficial societal outcomes.
• Clanker: A Word For The Machine — Using precise, non-anthropomorphic language for AI systems is crucial to maintain human responsibility and avoid blurring ethical boundaries.
• Joel Pearson on putting human first, 5 rules for intuition, AI for mental imagery, and cognitive upsizing (AC Ep25) — Effective AI integration requires human-centered design and psychological toolkits for managing change and augmenting human capabilities.
• Investigating Concept Alignment Using Implausible Category Members — AI systems exhibit significant concept misalignment with humans, particularly in category boundary judgments, leading to safety risks.
• Agents Are Not Enough: The Next Bottleneck Is the Human Framework — Human frameworks, not agent autonomy, are the next critical bottleneck for effective and accountable AI systems.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by cs.CL updates on arXiv.org.