Work Diaries are our virtual confession booth for people with jobs of all kinds to share their juicy—and anonymous—recounts exposing the good, bad and ugly of work in real life. Interested in submitting an anonymous Work Diary? Contact editorial@brightandearly.ca with the subject line ‘Work Diary’.
Role: VP, Not-for-Profit
Size of Company: 300-500
Total Years Working: 15+
Age: 40’s
When I first joined my organization two years ago, I had a really good relationship with our CEO. We had a lot in common, including similar professional backgrounds – she told me she saw a lot of herself in me. She was clear early on about how she liked things done and what her expectations were, but at the same time, she was hands-off and happy to let me run with the job as I saw fit. We worked well together.
Things started to change after she discovered ChatGPT.
Last year, she took a two month sabbatical to focus on herself. When she got back, things felt different. She seemed checked out, spent less time with me, and started relying heavily on generative AI for all of her work.
At first, her GPT use seemed pretty normal. She used it to summarize meeting notes and generate action items, helpful stuff. But over time, ChatGPT started to write, think, and decide essentially everything for her.
My team routinely gets multi-page documents from her, sometimes 30 to 80 pages long, that are supposed to be detailed plans for executing different projects. They’re clearly written entirely by ChatGPT; I can tell because she often leaves the prompts in. There’s context that isn’t accurate, departments listed that we don’t even have, and people’s names are often spelled incorrectly. Spelling people’s names seems like a small thing, but it tells me she hasn’t personally read it, and is overall a terrible experience for the team. She’ll also generate follow-up documents for these projects that contradict those that were generated before, leading to mass confusion. When we try to flag it, she tells us she doesn’t have time to review everything, and it’s up to us to identify issues and work through them. But how do I flag an issue when none of the documents make sense?
Since leaning into AI, she’s also limited meetings where we can have open discussions. Most of them are just reviewing project boards (also written by ChatGPT) full of tasks that don’t connect or have context. Then, she has ChatGPT summarize the meeting notes and send them out to us.
When we do have to communicate live, she’s even outsourcing that to ChatGPT. She literally reads from a script it generates! If someone asks a question, she’s generally dismissive and says she’ll get to it later. I’ve started wondering if it’s because she doesn’t have an answer prepared by the AI.
She also uses it to write all her messages on Teams. We’re a remote office and rarely in person, so online chat is supposed to be the place where we’re a little bit more fun and informal. Before her trip, she’d be a part of that. Now, we get these weirdly long, laborious sentences, with too many words in them, that I can tell are not written by a real person, let alone the boss I knew. ChatGPT is even being used for our performance reviews, which is a bit disheartening. I have a colleague who sent in questions about her performance review and received a response clearly crafted in (you guessed it): ChatGPT. She doesn’t seem to see the issue.
It’s really demotivating to have ChatGPT be your boss, and have ChatGPT be what reviews your documents, gives you feedback, and writes your performance review. If it’s doing the thinking, the planning, and the drafting, then I’m basically just a project manager trying to make sure ChatGPT’s vision gets executed. It makes me question what role is actually left for me.
I’m also concerned about integrity. If these documents aren’t even being read by a human, and that’s what we’re running on, who’s looking out for errors? Who is going to ask the important questions that a robot doesn’t know how to ask?
At a certain point, I don’t know how I can work long-term for an organization that feels like it’s being run by artificial intelligence.
