I suppose if you are being pressured to use AI at work, one thing you could do is send HR articles/studies on AI psychosis & ask "does [company name] accept liability for any mental harm that may occur from AI use?"

Before you touch any of the AI tools they want you to use, ask them to put in writing that the company is aware of the concerns about mental harm from AI use but is asking you to use it anyway. Ask to put a note in your employee file saying that you objected to using it.

Obviously being able to do this depends on a certain amount of job security.

But if you are fairly secure in your position, when you are told to do something shady at work, asking the company to put it in writing will often make them suddenly decide they don't actually need you to do that thing.

Obviously they have no problem writing "we want you to use AI." What they probably don't want to put in writing is "we are aware of the potential harm to our employees from AI use."

It's not magical, but "could I get that in writing?" can be surprisingly powerful.

You do have to be persistent & leverage documented communication as much as possible to get results.

Communicate over email if at all possible, & if you have a phone call or an in-person meeting about it, take notes & then send them in an email to all persons involved. Every single little convo you have about it, email the person & say "thanks for the talk, here's the content of our conversation."

Basically, get their refusal to put something in writing in writing.

Be polite as possible about it, no anger, no confrontation, just a friendly email with meeting notes & "here's what I heard you say, did I get that right?"

The trick with this type of thing is being entirely "by the book." Just communicate your perfectly reasonable belief that they will OF COURSE be willing to put anything they ask you to do in writing because OF COURSE it is on the level & OF COURSE they would never dodge responsibility for it.

Maybe act a little baffled if they don't want to commit something to an email or a memo. No accusation of anything just, "I don't understand, if this is how it is, why won't you put that in an email?"

Send that "baffled" response as an email. What you're doing is creating a paper trail. You can do that even if THEY refuse to.

And all the time, write in the most reasonable & friendly tone you can. You aren't upset: you are playing the role of a confused employee who doesn't understand why this is so difficult. They'll probably know that's not how you really feel, but the paper trail should show you asking for very reasonable things that no one can object to (without seeming shady).

It is key to maintain your confused but trusting persona, as frustrating as it may be. Write as though you truly believe that everything they do is on the level & that they always have your best interests at heart.

In this matter, you are just a pure of heart individual trying to get a little clarification on an issue that concerns you, & you would just like to have that clarification in writing. Since obviously this all on the level, there is no reason why they wouldn't do that for you.

@artemis "the only way this would be a problem is if they were acting in bad faith, so of course what you're doing is in good faith" is an important stance to learn and understand, yeah

including for being able to look the fuckers in the eye afterwards when oops, your entirely reasonable operating assumption totally fucked them over

(reasonably normal work isn't the most fucked up place I've used this)

@artemis
As a frequently confused-but-trusting individual, and one formerly pure of heart, can confirm.

@artemis This was pretty much my CYA technique at my last job.

Email, "Just wanted to confirm that <task you mentioned> was my number one priority for this week right?"

Next team meeting, "Why aren't you working on <other task>?, Well I got an email from <name> that said I should focus on this other thing, I CC'd you in on it. Happy to change but you'll need to talk to <name>."

@artemis dance like no ones watching
document like it’s going to be read out loud in a courtroom

@artemis

Hey, look, you're a valuable employee, but guess what, this economy sucks and we'll hafta let you go.
So sad.

@artemis this is easier said than done. There might be a few individuals pulling this off, the majority will not be able to. Depending on 1. country 2. Law 3. mngmt position 4. sector 5. market situation it will be difficult to get the employer to sign anything. In any case 1 will have to provide evidence/recognized studies/sources. Ppl depending on the job for their livelihood can be pressured/fired/replaced (points 3+ 5).

@Ilka4You @artemis 6. Autism
I try the friendly but confused thing all the damn time and it's frequently interpreted as hostile and I rarely understand why

That said, I do think anyone who thinks they can possibly get away with following this advice should. I would still try, if my job was asking me to use AI

@raphaelmorgan @artemis yes, individual situations can make it difficult to stand up against any sort of pressure or even abuse. It could be "simple things" like requiring flexibility to get a kid to school - the employer can hold it against you "I do so much for you - now do xy for me". Often one does not have the luxury to choose a fair employer.
@artemis There is a substantial body of proof that a lot of things happening in AI is not done in good faith. As it is marketed, why it is marketed, what the hosts of said AI gain in exchange for what the user does etc. Completely on board.

But AI Psychosis and pleading it to HR? I think the strike rate would be as high as pleading allergy to electricity. Unless there are some sources out there I have yet to read.
@artemis But make sure the articles are AI generated. Win win.
@artemis
Thank you. I'm saving your advice. So far my employer has only hired a full-time AI booster who tells us how awesome copilot is and provided online training. They continue the weekly reinstallation of copilot on my computer, which I promptly uninstall and then copy some more incomprehensible shit to my onedrive just in case microslop is scraping it.
@artemis Is there any connection to chatbot use _at work_ and the psychosis effects?
@artemis can you point me to such an article? The only ones I've been able to find are about people who try to treat their (pre-existing) mental illness/issues by talking to a chatbot, that is not what my employer is asking me to do. I've not been able to find articles that link AI use for office tasks to mental issues
@artemis I'd ask the employer to accept all liability and indemnify me for *any* harm to any person or legal entity caused by the use of AI in my employment. If they tried to brush me off, I'd call my union.