We Need to Talk About How We Talk About ‘AI’

https://www.techpolicy.press/we-need-to-talk-about-how-we-talk-about-ai

Publish Date: 2026-01-07 09:20:00

Source Domain: www.techpolicy.press

Here’s an unordered list summarizing the key points from the article:

Misleading Anthropomorphism: Anthropomorphizing language around AI technologies—attributing human qualities and abilities to them—masks important technical limitations and misleads the public’s understanding.
Risk of Over-reliance: Such descriptions promote misplaced trust and over-reliance on AI systems, leading users to attribute cognitive and communicative abilities to them that they might not possess.
Accountability Issues: The anthropomorphizing language obscures accountability by positioning AI systems as quasi-independent agents rather than tools used by people.
Historical Critique: The issue of using misleading language has been noted as far back as 1976 by computer scientist Drew McDermott, who warned against “wishful mnemonics.”
Alternative Communication: To avoid misleading interpretations, we should describe AI systems more accurately, specifying their functionalities without attributing human qualities or skills.
Impact on Vulnerable Populations: Misleading language can have a disproportionate impact on less informed or vulnerable populations, potentially leading to harmful interactions with AI systems.
Educating the Public: Improving “AI” literacy through more accurate language helps people make informed decisions about technology acceptance, contrary to claims that education might reduce adaptivity.
Collaborative Effort: All entities using or discussing AI—scientists, journalists, and the public—have a role to play in adopting clearer, more accurate communication about AI systems, thus fostering better public understanding and usage.