Opinion Essay Calls AI a Bad First Draft, Noting Limits Writers See

2 articles · Updated · The Washington Post · May 22

A May 13 reader essay argues AI functions like a “first draft” — and a bad one — because it can only mirror the material users feed it.
Herbert Lin’s earlier comparison of chatbots to palm readers underpins that critique: the systems reflect people back to themselves rather than generate genuinely original writing.
The piece says that weakness is especially visible to writers, who are trained to detect and revise flawed early drafts in ways engineers may overlook.
The argument extends a broader Washington Post opinion theme this week that AI is degrading trust and quality online, after another columnist warned of AI fakery and unreliable digital content.

As AI floods the internet with 'fakery', what is the future of trust and authenticity online?

Will new copyright rules for AI training data protect creators or just stifle technological innovation?