Use of AI in newspaper articles

arbitron

Registered User
Messages
795
A ‘cold, spreadsheet-driven’ Aer Lingus response to a young woman’s ‘freak accident’

This recent article by Conor Pope in the Irish Times has multiple quotes from the reader/consumer involved. Much of what she is quoted as saying is clearly AI.

It was probably a written submission by the reader, rather than a verbal interview. I assume she wrote her piece or an outline and asked ChatGPT to write or polish it. It is really off-putting to read that flat, robotic register.

The Irish Times has strict rules about staff using LLMs but interviewees are not restricted in their AI use. Some people struggle with literacy and find LLMs helpful, so it's not fair to ban it. But they could ask contributors to declare whether or not they have used AI and it could be noted in the article. At a minimum the reporting/editorial staff should pick up on this and address it in some way.
 
Paywalled so don't have access, but my AI was unable to discern any telltale AI hallmarks :)

On declaring use, there are arguments for and against and some of the models introduced various methods of watermarking, but there isn't yet consensus on whether that's a good thing or not. Most waternarking methods are being defeated shortly after launch. I think we're approaching a time where most content will pass through an AI of some form prior to publication or transmission.
 
I don't know if it was AI generated, or some attempt was being made to put it into 'business speak' so that corporate heads might get the message.
 
AI was unable to discern any telltale AI hallmarks
They are covering for each other! :p

The other issue is that LLMs are undoubtedly influencing human writing styles and even speech - some things that sound like AI are in fact just people who have picked up the AI speak.
 
  • Like
Reactions: Leo
Back
Top