Content-type: text/html Downes.ca ~ Stephen's Web ~ It's impossible to detect LLM-created text

Stephen Downes

Knowledge, Learning, Community

Despite the claims of (mostly edtech) companies, the evidence suggests that it is impossible to detect text created by a large learning model (LLM). "There will no doubt be an on-going LLM generation/detection arms race," writes Mark Liberman, "though it seems that the detectors have been losing badly from the beginning." Image: AI Weirdness.

Today: 1 Total: 1457 [Direct link] [Share]


Stephen Downes Stephen Downes, Casselman, Canada
stephen@downes.ca

Copyright 2024
Last Updated: May 26, 2024 5:34 p.m.

Canadian Flag Creative Commons License.

Force:yes