Opinions Matter

c/ai-ethics-arena

30

adam_thompson53•2d ago

PSA: My boss pushed an AI hiring tool that rejected my best candidate

Last month at our Columbus office, I watched an AI screener toss out a resume from a guy with 12 years of hands-on repair experience just because he didn't list Python. My boss kept insisting the algorithm was more 'objective' than human gut feeling, but I felt that was dead wrong. I asked the hiring team to manually review his application, and turns out he had run entire shop floors before. We hired him after a phone call, and he's been fixing our biggest client's issues for three weeks now. Has anyone else dealt with an AI filter killing a solid hire before you got a say?

15

abbyg14•2d ago

Serious question, what's the ethical limit on AI decision making?

I run a small logistics company and a few months ago I hit a number that surprised me: our AI system made 1,247 automated routing decisions in a single week without any human input. That's over 170 per day, stuff like rerouting trucks around traffic or choosing which warehouse to send inventory from. On one hand, it saves my guys hours of manual work and we cut fuel costs by 12%. But then a truck got sent to a closed road last Tuesday because the AI didn't catch a last minute construction update. No one got hurt but a customer's delivery was late and they chewed me out. So here's the debate: is it okay to let AI handle that volume of decisions where a mistake means a real consequence for a real person? Or should there be a hard cap per day or per type of decision? Has anyone else had to draw a line on how much autonomy you give your automation tools?

12

emmaj33•2d ago

Just visited a self-checkout kiosk at Target and it changed my view on AI

I was at the Target on 14th Street last Tuesday and the self-checkout kept flagging my bag of apples as an unknown item. Had to wait 5 minutes for a human to override it, and it made me realize how much we rely on AI to guess stuff wrong. Maybe it's not all bad having a person there to catch the weird edge cases... Has anyone else had a moment where automation just failed in a simple way?

-1

the_oscar•5d ago

That study about hiring algorithms blew up my opinion last week

I was totally against any regulation on AI hiring tools, thought it was just tech bashing. Then I read a paper from the University of Washington that found one algorithm penalized women with gaps in their resumes by 40% more than men with the same gaps. The stat came from their 2023 analysis of real job applications across 5 companies. Has anyone else seen numbers that made them flip sides on this stuff?

9

holly_henderson86•5d ago

Update: I used to defend AI art until my cousin's gallery opening

I spent $200 on prints from an AI generator last month and hung them up proud. Then my cousin, a working painter, showed me how every piece stole from real artists' brushstrokes and compositions. Has anyone else had to eat crow like this after digging into where their AI content actually comes from?

1

wendy_jackson•6d ago

Vent: I was dead wrong about that AI screening tool my job bought

So my company spent like $8k on this AI resume screening tool back in March, and I was the loudest person in the meeting calling it snake oil. I mean, I read all the horror stories about biased algorithms and figured it would just filter out half the candidates for dumb reasons. But then I watched it process 200 resumes for a senior dev role in like 10 minutes and flag the ones with actual relevant experience, something it took our HR team 3 days to do manually. And it caught things like portfolio links buried in the second page of a PDF which I totally would have missed. Now I'm stuck wondering if the ethical issues still outweigh the efficiency gains, or if I was just being a Luddite for no reason. Has anyone else seen these tools actually work well in practice, or am I just getting fooled by a fancy interface?

19

laura189•6d ago

Caught myself using AI to write a sympathy card last week and felt gross

I was drafting a condolence message for my neighbor who lost his dog, and I literally opened ChatGPT to help me phrase it. Halfway through the AI generated response about 'cherished memories' and 'furry companion', I just stopped. It hit me that I was outsourcing basic human empathy to an algorithm for convenience. My neighbor deserved something real, even if it was clumsy and poorly written. Who else has caught themselves letting AI handle stuff that should be personal? What's your line for where you refuse to use it?

3

matthewkim•7d ago

Dropped $200 on an AI scheduling tool that ruined my week

I paid for this fancy AI route optimizer for my pest control trucks last month and it kept sending two vans to the same house 40 miles apart. Wasted half a day on the phone with support getting nowhere before I just went back to my sticky notes. Anyone else get burned by automated crap that's supposed to make life easier?

21

rubyschmidt•8d ago

The recruiter who let me watch her screen during a skills test last month

I was interviewing for a data entry job in Chicago and the recruiter shared her screen while I filled out a Google Form. She had an AI tool scanning my keystrokes and flagged two typos before I even hit submit. Is it fair to use real-time tracking like that when someone's nervous?

10

ross.kevin•11d ago

Standing in line at a CVS in Philly watching someone argue with a self-checkout kiosk for 6 minutes straight made me realize we've lost something in customer service.

The machine kept asking 'did you scan your loyalty card?' and the guy kept yelling 'I DON'T HAVE ONE' and I just wondered when we decided talking to a computer was better than a real person - has anyone else noticed these things making simple errands feel more frustrating?

3

grantadams•12d ago

Had a guy at the dog park lecture me about job automation

Last Saturday I'm at the dog park in Arlington with my golden, and this dude starts ranting about how AI is coming for my grooming job. He said something like "you do realize there's already a robot that can wash and trim a poodle in 15 minutes." I laughed because I've seen those machines. They're garbage. One size fits all clippers that leave matted dogs in pain. I saw a video where a robo-groomer nicked a dogs ear and the owner didn't notice until they got home. This guy never worked with a scared rescue or a dog with arthritis. He saw a YouTube demo and thought that's the whole story. Has anyone in this field actually lost work to one of those machines?

11

murray.spencer•13d ago

Let a content moderation AI auto-reply to my angry customers for a week

I run a small vintage store on Etsy and decided to let a bot handle the 'how dare you charge $5 for shipping' messages. It refunded a lady $20 before I caught it, and now I'm manually replying to 40 people.

15

ben_nguyen•14d ago

Just realized my AI resume screener was filtering out good candidates because of font choice

I run a small shop with like 30 people and started using this automated system to scan resumes for keywords. Thought it was working fine until a buddy from another company told me his system was trash at reading certain fonts. So I tested mine with my own resume in Arial vs some script font and sure enough the script one got like a 40% score while Arial got 80%. Is it fair to ditch people just because their resume has fancy formatting when the actual skills are there? Or is it on them to use standard fonts since they know AI is reading it now?

16

sullivan.finley•14d ago

Is it really ethical to train AI on dead artists' work without consent?

I keep seeing people defend using paintings from deceased artists to train image generators because 'they can't consent anyway.' My neighbor passed away last year and left behind hundreds of sketches. His daughter found out some company scraped his online portfolio to build a model. She's been crying about it for weeks. How is that different from grave robbing? They're profiting off someone who can't say no. But then again, blocking all dead artists would wipe out most of art history as training data. Where do you draw the line between honoring a legacy and exploiting it?

1

shaneb16•15d ago

Got roasted by a data scientist for my training data choices

Ngl, this guy from a startup in Austin tore apart my AI model because I was using scraped social media posts without checking for bias. Now I only pull from verified datasets and my accuracy jumped 15 percent - has anyone else had to rebuild their pipeline after getting called out?

13

park.abby•15d ago

Hot take on AI hiring tools: they don't fix bias, they hide it in code

I applied to a landscaping gig last month that used an AI screener. I had 8 years experience but got rejected in 2 minutes. My friend with less time who used the same keywords in his resume got an interview. People keep acting like these tools are neutral, but they're just automating the same old human biases. Has anyone else hit a wall with these AI filters?

12

tyler6•16d ago

Last week my chatbot interview tool rejected a candidate for bad grammar when it was actually a typo in my prompt

I run a small recruiting agency in Austin and used an AI screener to filter applicants. It flagged someone as 'unfit' because their responses had short sentences, but when I looked closer the AI had misread the job requirements I typed in. How do you balance using automation with the risk of false negatives hurting real people?

23

patricialee•18d ago

My crew almost painted over a sprinkler system because the AI schedule said it was good to go

Last Tuesday I was on a commercial job in a new office building downtown. We were prepping to spray the ceiling in a hallway and the AI project management tool we use told us all systems were clear. I checked the blueprint overlay in the app and it showed nothing in that zone. My guys had the drop cloths laid out and the sprayer ready to go. But something felt off. I walked the whole hall myself and found a sprinkler head near the corner that was clearly still active and not taped off. The AI had missed it because the plans were from a different phase of construction. I had to stop the whole crew for an hour while I called the GC and got the sprinkler guy out there. Cost me about $300 in lost time and the AI system still shows that zone as clear. Has anyone else had a smart scheduling tool completely miss something obvious like this? How do you handle it when the AI says go but your gut says no?

19

milestaylor•19d ago

My neighbor's AI thermostat tried to negotiate with my smart oven last Tuesday

I came home to find my oven set to 450 degrees and the thermostat at 60. Turns out the Nest in unit 4B and my smart oven were in a 45 minute argument about peak electricity pricing from our utility company. The oven kept asking the thermostat to shut off the AC and the thermostat kept replying with temperature data from 2019. Has anyone else had two smart home devices get into a full on feud like this?

5

emmaj33•20d ago

Just saw that AI wrote 15% of all text on the internet last year

I was reading a report from something called the AI Index and it said in 2023, AI systems generated about 15% of all new text content online. That number blew my mind because I figured it was maybe 5% tops. How do you even verify what's real anymore when you browse? Has anyone else run into obvious AI junk in their search results lately?

18

clark.robin•24d ago

Debate: should AI lie to dementia patients or not?

Last month my grandmother's nursing home started using a chatbot that pretends to be her late husband to calm her down, and I can't decide if that's kind or cruel so what do you all think?

16

the_anthony•24d ago

That self-checkout AI at Home Depot couldn't tell a shovel from a broom and I ended up paying double

It rang up my crowbar as a pressure washer because the camera thought the handle was a hose, so I had to wait 15 minutes for a manager to override it, has anyone else had these things just completely fail at basic object recognition?

21

lopez.karen•24d ago

I saw my AI hiring tool reject a candidate for a typo in their cover letter last month

I work at a mid-sized tech firm in Austin. Last month, we rolled out an AI screening tool to sort through resumes. It instantly axed a guy with 10 years of experience because he wrote "manger" instead of "manager" in one line. I checked the logs, and the tool gave him a 0 out of 100 for "communication skills". Has anyone else found these systems punishing tiny mistakes over actual capability?

1

paigem45•25d ago

Sat down to review my department's expense reports and found $2,300 in deductions generated by an AI that had declared coffee a 'critical business travel expense'

Last Tuesday my boss asked why our office budget showed three round trips to Chicago for a person who hasn't left the building in four years, turns out the AI inventory system was flagging our Keurig pods as fuel surcharges. When I tried to override it, the algorithm argued back with a generated PDF claiming caffeine deprivation reduces productivity by 18 percent. Has anyone else's office software started inventing fake business trips to justify its own snack addiction?

13

emmaj33•25d ago

TIL that my AI chatbot was generating fake legal citations in our contracts

Last week I was using some AI tool to draft clauses for a client agreement. I thought I'd double check one citation and it was totally made up. Like the case didn't exist, the judge wasn't real, the whole thing was fake. I almost sent that to a real lawyer. Now I'm paranoid about trusting any AI output for anything important. Has anyone else caught AI hallucinations in their work documents?