Urgent.News

What's breaking now, across hundreds of outlets.

Tech

Human, for Now: The Adjective Is Changing Sides

Two years ago that trailer was the interesting part. It flagged the unusual thing. I think we're close to the point where the interesting line is the other one: what did the human do? Cambridge Dictionary seems to agree. On 17 November 2026 it named human its Word of the Year. Lexicographer Colin McIntosh's reason: "AI is changing how we talk about being human." ( announcement ) The adjective is…

In the past two years, that trailer was the unusual aspect that stood out. However, the intriguing focus is shifting to the human actions involved. Cambridge Dictionary recently designated "human" as its Word of the Year, reflecting the belief that artificial intelligence (AI) is transforming our perception of what it means to be human.

This change is evident in the evolution of terminology, with "human" being used to qualify AI-generated, AI-written, AI-edited, and AI-created outputs. As the exception becomes more common, the ordinary is taking on exceptional characteristics, leading to the labeling of humans as "human-written," "human-created," "human-edited," and "human-reviewed."

Linguists refer to this phenomenon as markedness, where the default is unmarked and the exception carries a label. When the default changes, the label moves accordingly. In the past, no label was needed, but now, as AI-generated content becomes more prevalent, the need for labels is growing in more places, from "Human Authored" marks for books to discussions around "human reserved jobs."

The adjective "human" is being asked to do a more challenging job than "handwritten" ever had. It carries four claims: human-authored (a person wrote the substance), human-reviewed (a person checked it), human-in-the-loop (a person has a step in the process), and human-accountable (a named person answers for the work). Developers have observed the gap between AI-written content and human involvement.

While provenance systems like Linux's "Signed-off-by" and "Reviewed-by" have been established for code, the human side lacks the same level of precision. A proposed solution is to include a "human" field in the code commit metadata, specifying the role played by AI and human authors, as well as details about edits and reviews. This approach could help distinguish between partial and full human involvement in various artifacts, such as code, documentation, reports, and generated summaries.

The Authors Guild's "Human Authored" mark for books serves as a good model, as it requires registration and acceptance of a license, and includes a de minimis exception for spelling and grammar tools. However, the proposed system would require more structure, such as threshold levels, registration, and named exceptions. The ultimate goal is to find a useful label that captures the nuanced roles of humans and AI in content creation, moving beyond simple adjectives to a more comprehensive metadata system.

Rewritten by urgent.news; may contain errors — check the original.

Source: dev.to

Read the original at dev.to →

Report a problem with this story

More in Tech

What is an agent execution receipt standard?

The short answer An agent execution receipt standard is a published specification that defines, for a record of one AI agent tool call, four things exactly: what the record contains, how the…

  • Agent execution receipt standard defines essential details of AI agent tool calls
  • AER-1 is a published IETF Internet-Draft exemplifying the standard
  • Receipt record includes execution details, commitment, verifier outcomes, and conformance corpus

More from Sunday 22 November →