The Parrot in the Machine
Emily Bender and Alex Hanna argue that large language models are not intelligent but stochastic parrots mimicking language without understanding, and that AI hype distracts from real harms like labor exploitation and bias reinforcement.
Background
- Emily M. Bender is a computational linguist at the University of Washington and a leading critic of AI hype, known for co-authoring the influential 2021 paper "On the Dangers of Stochastic Parrots" (which led to the firing of a Google AI ethics researcher).
- The piece reviews Bender's book *The AI Con*, co-authored with Chirag Shah, which argues that large language models (like ChatGPT) are not "intelligent" but sophisticated pattern-matching machines — "stochastic parrots" that predict plausible words without understanding meaning.
- Bender and Shah warn that corporations (OpenAI, Google, Microsoft, Meta) are aggressively marketing these systems as sentient or superhuman to extract investment and shape policy, while obscuring their flaws: unreliability, bias, environmental cost, and enabling surveillance or misinformation.
- The review examines how the book connects technical explanation to political critique, rebutting claims of "artificial general intelligence" (AGI) and arguing that AI discourse functions as a con — a deliberate framing to concentrate power and wealth.