Modern Editorial Ethics Question

[howard lee]

Hypothetically speaking, let's say you are one of the editors of an online blog for a major retailer of imaging and tech gear. You have worked your butt off for 14.5 years to craft a corporate voice, clean up grammar and syntax, strive for quality and return readers, and have taken a great deal of pride in your work and the fact that you have been largely responsible for the quality of the content.

One of your writers brings to your attention the idea that the copy of two of the recently contracted freelancers (whom none of you have met, but have some presence on the Internet) seems stilted and lifeless, and she believes management has introduced AI-created copy into your production workflow without advising the editors or anyone else. And let's say, hypothetically, that you have copied and pasted some of this suspect copy into a Web tool that can identify AI-originated copy, and furthermore, the result is "AI Text Detected."

How do you feel about this? Do you find it to be a breach of ethics, or a fraud? Would you be bothered by it? Angered? Insulted? Outraged? Do you say something to upper management (most of whom wouldn't recognize excellent prose if it bit them on the ass), or do you just keep your head down and do as much rewriting on the QT as you can to introduce some soul and verve into this otherwise dreadful, Sahara-desert-dry copy?

How would you handle this?

I look forward to reading your opinions. Thanks in advance for taking the time.

[majorminor]

How close to retirement are you?

[dradtke]

I would ask myself, "How close to retirement am I?"

[howard lee]

I'm not sure I want that information on the Internet right now... besides, this is a hypothetical question.  😉

[Cornflake]

"How do you feel about this? Do you find it to be a breach of ethics, or a fraud? Would you be bothered by it? Angered? Insulted? Outraged? Do you say something to upper management (most of whom wouldn't recognize excellent prose if it bit them on the ass), or do you just keep your head down and do as much rewriting on the QT as you can to introduce some soul and verve into this otherwise dreadful, Sahara-desert-dry copy?"

For starters, are you sure it wasn't the freelancers who introduced the AI text? It seems improbable that two of them would do it at the same time but it's not impossible. How confident are you that your web tool accurately detected the presence of AI?

If it happened to me, I'd cool down before doing anything. I know I don't make very good decisions when I'm angry.

"Do you find it to be a breach of ethics, or a fraud?" No. Unless I've missed something, management can do pretty much anything it wants with with its blog. It could fire all the humans and let the robots run the whole thing. (I wouldn't recommend it.) Bill or Russell may know of some ethical constraints that I'm not aware of.

"Would you be bothered by it? Angered? Insulted? Outraged?" Yes. What's wrong with this hypothetical scenario is that management didn't let its employees know what it was doing. My guess is that it didn't have any legal obligation to let them know. But if you want employees to be loyal and do their best work you have to show them some respect. That wasn't done here.

I don't know all the ins and outs of the situation but my inclination would be to stick to the chain of command. I'd explain my concern to my supervisor, ask whether AI text was being used and ask why nobody was told about it. There's always the possibility that there's a good explanation.

[howard lee]

Hypothetically speaking, there's no question it was AI generated, Don. Not only is the Web tool is dependable—the parties involved tested it against some known human-generated copy and it was able to identify that accurately—but an experienced writer/editor can tell just by reading it; it's dead, like a shark's eyes. I think the supervisor you suggested speaking with may, hypothetically, be party to this hypothetical conspiracy and it would probably not be a wise move, hypothetically, to approach him. He has always been a milquetoast, was promoted because he will always say yes to management, and has recently seemed to grow a set of cojones, at least, in text messages. In this hypothetical situation, it's a red flag to me.

You have made some very good points. Thank you.

[amanajoe]

Hypothetically, don't trust AI detection tools.

www.technologyreview.com/2023/07/07/1075982/ai-text-detection-tools-are-really-easy-to-fool/

While they aren't useless, many people are being accused of using AI generated text (false positives) that have never even thought about AI, let alone used it. If you'd like to adjust your confirmation bias, take something you wrote years before AI and send it through the detection tool and see what it has to say about your own work.

[amanajoe]

Nov 18, 2023 11:16:41 GMT -5 howard lee said:

Hypothetically speaking, there's no question it was AI generated, Don. Not only is the Web tool is dependable—the parties involved tested it against some known human-generated copy and it was able to identify that accurately—but an experienced writer/editor can tell just by reading it; it's dead, like a shark's eyes. I think the supervisor you suggested speaking with may, hypothetically, be party to this hypothetical conspiracy and it would probably not be a wise move, hypothetically, to approach him. He has always been a milquetoast, was promoted because he will always say yes to management, and has recently seemed to grow a set of cojones, at least, in text messages. In this hypothetical situation, it's a red flag to me.

You have made some very good points. Thank you.

You responded while I was typing. So, as an editor, do what editors do, send it back and say do better.

[howard lee]

Nov 18, 2023 11:21:36 GMT -5 amanajoe said:

Nov 18, 2023 11:16:41 GMT -5 howard lee said:

Hypothetically speaking, there's no question it was AI generated, Don. Not only is the Web tool is dependable—the parties involved tested it against some known human-generated copy and it was able to identify that accurately—but an experienced writer/editor can tell just by reading it; it's dead, like a shark's eyes. I think the supervisor you suggested speaking with may, hypothetically, be party to this hypothetical conspiracy and it would probably not be a wise move, hypothetically, to approach him. He has always been a milquetoast, was promoted because he will always say yes to management, and has recently seemed to grow a set of cojones, at least, in text messages. In this hypothetical situation, it's a red flag to me.

You have made some very good points. Thank you.

You responded while I was typing. So, as an editor, do what editors do, send it back and say do better.

howard lee LOVES this.

[epaul]

Hypothetically, how close to retirement is this hypothetical person?

Also, if they want AI, let them have it in all its glory and see what happens. Don't waste your time trying to polish a turd. Your job is to edit, not re-write crap cheapsters have cheaped out for.

[John B]

Nov 18, 2023 11:23:34 GMT -5 howard lee said:

Nov 18, 2023 11:21:36 GMT -5 amanajoe said:

You responded while I was typing. So, as an editor, do what editors do, send it back and say do better.

howard lee LOVES this.

"Your writing is dead, like shark eyes. It's hard to tell a human was even involved in this."

[drlj]

I am retired. I wouldn’t even look at it.
I might ask Alexa what she thought, but that is it.
Big yawn detected.

[Russell Letson]

I think Don nailed the legal/organizational issues--matters of fraud, libel, or contractual agreement aside, a private company can pretty much do whatever it wants with copy that will be presented as "theirs."

I can imagine a couple of scenarios. First, actual freelancers using AI tools to generate the copy they're hired to produce. From what I've heard about the pay for "content" production, the incentives to employ such shortcuts would be enormous. The contracting managers might not know or care whether a human produced the copy--it is, after all, just grist for whatever mill they are operating. Second, there are no outside contractors--instead, someone in-house is turning the AI crank and introducing the product into the normal workflow. But then the question is, why bother with such a deception? (Actually, I can imagine a couple of scenarios where that might happen, but they involve significant social/personality pathologies, or maybe an internal scam.)

As to how an editor on the sharp end of this might react, that depends on the shop culture. In a healthy organization, sending a message up the chain of command would be reasonable. In a less-healthy one--where, say, that chain consists of incompetents or lackeys or shirt-tail relatives--that would be an exercise in futility. And going around the chain of command just means confronting the jerks at the top who set up that chain of fools.

C. has been seeing suspect student work all term, and while the experienced eye can spot it immediately, proving this kind of intellectual dishonesty is much more difficult than proving old-fashioned plagiarism. (And ironically, much AI-generated work reads like the kind of C-minus student writing that comes from kids with middle-school literacy skills.)

[Cornflake]

The more you tell us about the situation, the more I lean towards doing nothing.

[Cornflake]

Well, not nothing. Have a beer.

[drlj]

Copy additure is a innecessery job anywhey.

[howard lee]

Nov 18, 2023 11:29:24 GMT -5 epaul said:

Hypothetically, how close to retirement is this hypothetical person?

Also, if they want AI, let them have it in all its glory and see what happens. Don't waste your time trying to polish a turd. Your job is to edit, not re-write crap cheapsters have cheaped out for.

Hypothetically, four and a half months.

You also make a very good point, Paul. The more I think about it, the more it's clear that if hypothetical managers don't care about it, then why should the hypothetical editor? Aside from all his years of labor, he has no hypothetical pigskin in this game.

Hypothetically speaking, the best thing to do is what Cornflake recommends: nothing.

[Marty]

Nov 18, 2023 12:24:11 GMT -5 howard lee said:

Nov 18, 2023 11:29:24 GMT -5 epaul said:

Hypothetically, how close to retirement is this hypothetical person?

Also, if they want AI, let them have it in all its glory and see what happens. Don't waste your time trying to polish a turd. Your job is to edit, not re-write crap cheapsters have cheaped out for.

Hypothetically, four and a half months.

You also make a very good point, Paul. The more I think about it, the more it's clear that if hypothetical managers don't care about it, then why should the hypothetical editor? Aside from all his years of labor, he has no hypothetical pigskin in this game.

Hypothetically speaking, the best thing to do is what Cornflake recommends: nothing.

Very much a Don't Rock the Boat situation.

Cheri retires in 6 1/2 months. There are quite a few issues at the Bank she would like to shout about but since she got grandfathered into a very nice retirement package and a Bank backed 401K she ain't gonna rock the boat. I married the right girl that stuck with the same company for 35 years. I will be a kept man.

Just let it go, do your time and walk away.

[Dub]

Marty gets my vote.

One of the rules I learned to live by during my decades in the military-industrial complex is never try to protect the company from itself. Even if you succeed, you will be forever branded and loathed. Let the company be as it wishes to be. Companies don’t have “ethics,” but they are bound by laws. Don’t involve yourself in breaking the law but everything else is up for grabs. Your goal is to extract the maximum amount of value (money, benefits, admiration, etc.) from your time there as you reasonably can. The rest is theirs.

[John B]

Nov 18, 2023 10:47:30 GMT -5 howard lee said:

Hypothetically speaking, let's say you are one of the editors of an online blog for a major retailer of imaging and tech gear. You have worked your butt off for 14.5 years to craft a corporate voice, clean up grammar and syntax, strive for quality and return readers, and have taken a great deal of pride in your work and the fact that you have been largely responsible for the quality of the content.

One of your writers brings to your attention the idea that the copy of two of the recently contracted freelancers (whom none of you have met, but have some presence on the Internet) seems stilted and lifeless, and she believes management has introduced AI-created copy into your production workflow without advising the editors or anyone else. And let's say, hypothetically, that you have copied and pasted some of this suspect copy into a Web tool that can identify AI-originated copy, and furthermore, the result is "AI Text Detected."

How do you feel about this? Do you find it to be a breach of ethics, or a fraud? Would you be bothered by it? Angered? Insulted? Outraged? Do you say something to upper management (most of whom wouldn't recognize excellent prose if it bit them on the ass), or do you just keep your head down and do as much rewriting on the QT as you can to introduce some soul and verve into this otherwise dreadful, Sahara-desert-dry copy?

How would you handle this?

I look forward to reading your opinions. Thanks in advance for taking the time.

I would be bothered, angered, insulted and outraged.

I might point it out to upper management. The way I would approach it would be along the lines of, "I don't know if you're aware, but I'm pretty sure the new freelancers you brought on are using AI to write their copy. If you are aware of it, that's fine, but if not you are way overpaying them for poorly-written, computer-generated content, and you should probably consider using other freelancers. And also, if the company policy is to start using AI to augment our content, I'd like to know about it. What I've seen so far is a big drop in quality from what our staff writes, and it makes my job harder. If I know beforehand what I'm editing is AI, I can adjust how I edit accordingly."

It may not make any difference, but maybe it will. It's not hand-wringing, "AI is the end of the world" but it lets them know that you know, and as I put it, you're more concerned with how it affects the company than any sort of moral argument for or against. And there is the chance they don't know, and the freelancers are making relative bank for using ChatGPT and the company is bankrolling it.

And frankly, you could propose to cut out a chunk of writers altogether, generate the copy yourself and then edit it. Saves the company money and keeps you employed for 6+ months, at least. Maybe even a consulting engagement/freelance position after hypothetical retirement.