Announcement
Humanlike language may not change how people view AI in a single session
A post about an updated paper says its study now includes 815 participants and offers stronger evidence that describing AI as “thinking” or “knowing” does not affect their views of it. The poster calls the evidence for no effect moderate.
TLDR
The person sharing the updated paper says its participant count doubled to 815, with stronger evidence that anthropomorphic language does not affect people's views of AI. They say the evidence is still moderate. The study tested English-speaking participants in the US during a single session, so a small effect could still build over time.
Combined views
2.2K
3 Sources, first seen 7h ago
