AI makes an appearance at my union meeting
I had an interesting experience this week. Not all union meetings are interesting, even if they are useful. Yesterday was a pleasant exception where it was both useful and interesting. For the first time, I witnessed AI coming up as a topic of conversation. There is no secret that people fear losing their jobs due to AI automation, and sure enough I saw proof of it to the extent that the union may consider adding some clauses around protecting jobs from AI.
How is it at your workplace? Where I work, this year I witnessed a very strong push to use AI. Messaging around using AI at town halls, messaging around using AI in team meetings, articles on the intranet site, IT events around how to craft good prompts, etc. I would not be surprised if they tied some leaders' bonuses to how much they can get their teams to use AI. This part is quite annoying to me, not to mention deceitful. If I were a leader I'd straight up tell my team about it. I am not a leader - leaders are not part of the union to begin with.
The whole thing made me also think about how my colleagues use AI. It really is a mixed bag. I see everything from the person who runs a 2-line email through AI five times to finetune every word, to myself who only reach for AI when I am stuck and it's just much faster than a search engine/forums/videos to solve my issues (for example needing a script in a programming language I am not familiar with).
I have AI fatigue but also use it constantly. I use it to work (software) because it shines there and makes me a lot faster. I use it personally to help me think of ideas, which I always feel a little guilty about.
My job definitely had the big push for AI adoption, but we're small enough that it was mostly just urging everyone to do it. Nothing specific. We're a startup, so it's practically due diligence to squeeze every new idea for value.
I'm definitely scared of losing my job to AI.
I feel guilty using it outside work, so I limit my use quite a bit.
Sucks to have that worry about AI taking your job. I hope it won't happen.
From my end, I don't think my job is threatened. There is too much work that is either an exception or needs a human touch. AI could automate a few things in my job. I guess a day may come when they may say "instead of 5 people on this team, we could make it work with 4 + AI", but I don't think it's anytime soon. Generally, even with all the AI encouragement, the organizations needs more people, not fewer.
It's being pushed very aggressively at my job (software). There's metrics on adoption and usage rates under each manager, which don't seem to be plugged into the IC performance review process but may be a part of the manager performance review.
I'm just now starting to find it useful, but it's nowhere near the panacea that the AI companies are trying to sell it as. I find the inline code completion stuff worse-than-useless (plain old LSP completion is far more useful) but command-line chats where it can edit files and run commands on my behalf are useful in limited circumstances.
My team's codebase is large and sprawling (over 300 repositories of varying size and importance) so there's a lot of cases where AI just can't understand what things are doing. When I'm doing something that has a lot of prior art (major dependency upgrades, for example), it's been relatively useful, but still requires a lot of guidance from me.
I, too, find it nowhere near what the hype would make one believe. I wish the whole thing was positioned differently, sold differently, even named differently to be honest.
I work for a small business, mostly we've talked about how it can be used for data entry, which is a good portion of the work. Haven't used much of it yet, but it looks like AI will add setting up AI, feeding the AI and AI babysitting in addition to the client babysitting we currently do - where we basically clean up after clients who constantly make a mess of their books rather than paying us to do it right the first time (which would cost them a lot less in the end).
We're in the business of helping people who are terrible at bookkeeping, mostly small businesses. I've seen no decrease in the number of people who would rather outsource thinking. AI is great and all, but if a business owner doesn't understand it then they won't be using it or will do a terrible job setting it up or choose a cheap AI and make such a horrible mess they'll need someone to sort it out.
I still have clients who can't handle logging into a simple file sharing service, which only requires they remember their username and password. They aren't even old! It's crazy how bad some of these people are outside their profession. As long as those types can run a business then we'll do fine.
Also, for us anyway, I feel like AI won't be graduating to the point it doesn't need constant assistance anytime soon, just a way to make our job easier or at least more productive. On par with asking a fast, amazingly focused 6 year old for help with data entry. ...I kept retyping that and reducing the age (can a 4 year old be taught to do data entry?)
I admit my exposure to AI data entry outside of QBO has so far been services clients found and those services must have dug their AI out of some hole under a burning dumpster on the internet. At least program them to question a receipt it thinks is a decade old! The one that created thousands of vendors was special, it would add a new vendor because it learned to ignore the end numbers on a bank transaction but only until a space or symbol so even though it's clearly the same vendor, a new vendor was created for nearly every transaction. Just random stuff like that happens and since it's automated it's ignored longer, like a rumba that automatically cleans the carpet in a house with pets (which at least used to get rather messy, if you were particularly unlucky).
AI babysitting is definitely required, and yes, as long as people will be bad at basic tech stuff AI won’t take over fully. So, never. My own experience with it was that it helped the very clean part of the manual data processing, but even small exceptions cannot be trusted. It reduced the time I spent but replaced one manual task (boring manual copy paste) with a different manual task (manually clean up all the exceptions AI can’t handle). Still, overall it did cut down an activity that took 8 hours to maybe 2 hours.
Thankfully my organization seems to be through the worst of it already.
Last year every time I went to a conference, every presenter thought they were being original and hilarious by having AI write something for them. 90% of team building exercises are still obviously AI generated. Thankfully there are only a few people who have full access for now, so I'm not saturated with AI emails and messages yet. Technically I'm pilot testing copilot integration with MS office... But I also technically haven't used it yet!
I have come to tell you it's pretty terrible. 90% of the time I ask it to do something, it can't do shit. Even when I don't ask it to perform an action and just throw a normal prompt at it, it does terribly, whereas normal copilot actually works pretty well.
I think I'm in the middle of it, at the height of the push. I look forward to the whole thing cooling down a bit. In the meanwhile I got so annoyed with Co-pilot integration in MS Office apps that I asked them to disable it, and leave me just the standalone co-pilot app so I can use it when I choose to rather than Clippy-style-in-your-face-every-second.
I'm in a public educational institution so the only threat our slow moving, people centered organization is seeing is by way of students and the instructors who have to grade their AI slop papers :(
Brutal… maybe you can use AI to grade AI. I joke, I joke. I spent some time in teaching, and remember grading was already the worst (most tedious) part of the job, and it was before AI. I can only imagine how “fun” it is now.
Well, R/Professors is where you can find your answers. It continued to be among my favorite subreddits.
We don't push it strongly, but it's encouraged to be a companion to your job. It is the biggest improvement for me personally for many years, and I try to help anyone that wants to learn more about it.
I lead our organization from a data perspective and we will have a lot more ML and AI used next year, where it makes sense. Our data isn't ready though.
That was my quick realization as I started using this to process data. I thought it would be much easier and powerful if the data were laid out in a better way. Not unique to AI stuff, but makes me think about that side of it for future improvements.