AI and current journalism standards

I started working for a B2B website in 2021 as a daily writer and reporter, it got sold in late 2022, but in early 2023 I was named editor. In the spring of this year the entire editorial team was eliminated and they moved to all AI content generation. We had a team of up to five writers, and we had developed a pretty solid publication and traffic was the highest it had been the entire time I was at the site when we were let go.

They paid pretty well, but the new owners really didn't know what they were doing in terms of business or media. I feel, although I don't know for sure, they were struggling financially and they saw AI as a way to do what they were doing without paying for it. They never talked with me about it, just did it. I suppose AI could work in this way with some effort but it still needs to be edited (the site is really a mess of style and everything else right now). One of the owners was especially fascinated by AI from the start, and for everything -- content, images, social media posts, whatever. I was always cold to it, so I think that's why they never at least addressed it with me.

Anyway, I am biased and I have mostly moved on, but I do still check in on the site. Like I said, it is a mess and I hate it because we had built such a solid platform. They recently started using bylines after three months or so of none. I believe these are not real people, they are super common names and there is no connection anywhere to this company -- the owners also love LinkedIn and a big red flag is no writers are connected to them currently. There is also nothing on the website about any writers (before there was a huge editorial landing page).

So are there any standards or ethics to any of this? Should a media source that presents itself as some sort of leader on the subject be clear to its readers that it is artificially produced? Is there a benefit to using fake people? I am guessing they got some pushback from advertisers and other users on their content and lack of people, but that is just a guess. Should they let people know that these are not real people (I have seen bylines elsewhere that have a name but indicate it is AI produced)?

I am getting kind of old in the industry and AI is not something I am well versed in. I am willing to learn, and do believe it can be a tool, but I don't think it replaces people in creativity or even total fact checking. One of my old writers, who works for a very respected publication, said he's had to sign contracts stating he will not use AI, so it isn't widely accepted yet.

But when used what should the public or clients be informed of? What is overall acceptable? Anything?

 

If you can’t be bothered to write it, why should I be bothered to read it?

 

Just a reminder that AP has been pumping out AI- generated gamers for years.

 

It’s a cliche I know but AI is coming for all jobs. The question is which humans will learn the fast how to tame it to their advantage.
Personally I am excited by it. As the owner of a marketing agency and a customer contact company I see huge advantages to it.
But @Roscablo to answer your original question, IMO AI or bot-generated content should be labeled as such. I think pretty soon consumers won’t care.

 

Anyone got any Tansa stories?

We're getting it soon. Weeeeeeeeeeeeeeeeeee, DOGGIE!

 

I think it can be a good tool. I have still seen too much in journalism where not at least having an editor can be a big problem (let's just say one time I had to deal with a Congressman's office for some AI content gone wrong). I also think in a content sense a vast majority of it right now is pretty obvious it is AI, and I do think that turns off users. That said, I also think users and readers in the end aren't going to care. However it ends up being used in media and journalism at this point I do wish it was labeled as such. You see it some, but not enough. Then let the reader totally decide.

 

Whenever I want to see how bad AI stuff can be, I just read a generated gamer on GameChanger. Well, I used to. Now you have subscribe and download an ap just to hopefully get a boxscore (Will it have full names? Will it at least have last names? Will it just be jersey numbers? Who knows! Tune in to find out.) and read a terrible writeup.

 

As the GameChanger person for my son's team, this is so true. I rarely look at them, but when I do they don't capture at all what actually impacted the game. It's almost like they have it set to get as many names, whatever format they may be in, in the story as possible.

 

They're hiring writers to help their machine language be, well, less machine-like. Interesting experiment.

 

That's just it, I don't think they hired writers. I think they are using fake names (or untraceable names because they are really common). Does this some how appease clients or even SEO situations?