I Am Begging AI Companies to Stop Naming Features After Human Processes

Anthropic announced a new feature called “dreaming” at its developer conference, describing an AI agent sorting through its “memories” to improve performance. This move highlights a trend where AI companies apply human cognitive terms to describe machine functions. While these metaphors can make AI more relatable, critic Jacob Steinhardt argues this practice creates confusion and exaggerates what AI systems actually do.

Naming AI features after human mental processes like dreaming or memory gives the false impression that machines possess consciousness or experience emotions. This risks misleading users, developers, and policymakers about AI’s capabilities and limitations. It blurs the line between metaphor and reality, potentially resulting in unrealistic expectations or misplaced trust in AI systems. The use of such labels also complicates technical understanding by masking straightforward computational mechanisms with poetic language.

The drive to anthropomorphize AI stems partly from marketing and partly from a natural desire to explain complex technology through familiar concepts. As AI systems increasingly incorporate memory storage or simulate human reasoning, companies label these processes with terms rooted in neurobiology or psychology. However, this tendency fails to capture the fundamentally different nature of AI cognition, which involves data patterns and algorithms rather than subjective experience.

This trend signals a broader challenge for the AI field: how to communicate new developments accurately without oversimplifying or overhyping. The industry must find a middle ground between dry technical jargon and misleading anthropomorphism. Clear language will help users and stakeholders better grasp what AI can realistically do and its limitations. Going forward, developers should aim for names that describe function rather than evoke human consciousness, prioritizing transparency over appeal.

We should watch whether this naming convention continues or if a pushback arises emphasizing clarity in AI communication. How companies describe their technology shapes public perception, regulatory approaches, and research directions. The next move might be more precise, less metaphor-driven explanations that respect the unique nature of artificial intelligence, supporting informed discussions and ethical development.

— AI Quick Briefs Editorial Desk