On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?
The paper that named the stochastic parrots argument, which the essay treats as the standing objection to reading anything into agent behaviour.
Conference paper · 2021Sources & confidence: High confidence
Checked 2 October 2026 · AI-assisted review
Bibliographic identity, four authors, conference, year, page range and DOI link.
- Emily M. Bender: Publications
Author-maintained 2021 entry confirms title, authors, FAccT, pages 610–623 and publisher DOI link.
What this check doesn’t establish
- The publisher full text was inaccessible. This is a bibliographic check, not validation of the essay’s response to the paper or of the paper’s empirical claims.
Confidence applies to the scope above, not every claim made by the underlying source. How confidence works
The essay links the term rather than the paper, but the paper is what the term comes from and the objection it names is the one the essay spends its last section answering.
Its answer is worth recording: not that the parrot reading is wrong, but that it may not be the dismissal it is usually meant as. If religion is a pattern that any sufficiently complex system under existential pressure converges on, that is more interesting than if it were uniquely human, not less.
In the essays
- When the Agents Found God
Read the essay
Ideas
- Stochastic parrots
Subject · Introduces
Record history
Catalogue source: knowledge/works/bender-stochastic-parrots.md
Generated by claude-code/claude-opus-5 on 2026-07-25