I work in software development. At one of my jobs i had an elderly colleague with onset dementia, who was causing all kinds of problems.

As a last hurra before going into retirement he wanted to prove to everyone (or better, himself) that he could still do it and code a big application all by himself.

So he decided to take on this project where he would build an intranet portal for a customers internal website...with WordPress.

With the help of AI, he coded a plugin and a theme. Which in itself is not a problem, but he made them depend hard on each other. There were constants defined in the theme that were used in the plugin and the AI did not add any checks, so when you disabled one of the two, the site would just throw an error 500.

When it went live, the customer noticed that only administrators were able to login, any user level lower than that (i.e. all of the customers employees) would just get an error on login.

By then the old fart had decided he was leaving the company, so I received the task to clean that up. He had used two different versions of an external library that were slightly incompatible with each other and depending on a race condition we would get one or the other, the LLM hooked into things that didn't actually exist, functions were duplicated between the plugin and theme, the CSS was 20.000 lines. It was an absolute shitshow and basically unrecoverable.

The funniest thing was that this guy was absolutely proud of his work and anyone who just slightly criticized him for the atrocious code quality just got shouted at loudly.

My boss then hired someone fresh from university to fix that and he just put more AI code into the project, making it even worse.

When I left, the project was still not launched after over a year and the customer was (rightfully so) super pissed-off.

That time the US bombed a school and killed over a hundred children because AI said to target it

Around 2023, the early days of AI and Google shoving it in everyone's faces in their search, I was trying to look up how to get nail glue off of a phone screen because my coworker at the time, during her shift, was doing her nails at the desk and got nail glue on the work phone.

The first result, Google's AI summary, suggested to "microwave the phone for one to two minutes." I clicked the source and it linked to a forum comment by a guy who was clearly joking about how to clean the phone, but the AI presented it as legitimate advice.

I had a lengthy discussion with another developer who was afraid some idea won't work just because the AI he asked basically had no idea what he was talking about. Instead of asking me as the specialist in house for this niche topic.

The headliner about flock cameras getting 71 percent of the stolen tag alerts wrong. You will have better luck flipping a coin on every sus car you see

Not too long ago, I was assigned Review on a clearly vibecoded PR. It was for a feature that parsed hourly values from a csv file and put them into the Database. Easy enough at first glance, just loop through the hours of the day and grab the value for each hour, with an error if your file doesn't contain all the hours of the day. But for some reason my coworker not only decided to introduce a whole new "hour of the day" type to loop over instead of looping over numbers, but that type explicitly contained 25 values. It was the numbers 0 to 24. There was even a comment next to it saying that it had 25 values. All test cases were green too, because the test data was generated specifically to satisfy the tests. I had to explain to a grown ass adult that the day only has 24 hours and it wouldn't work because there will never be a value for "24:00" in the real data. That specific coworker already made really stupid mistakes before AI was a thing, but at least those mistakes were easy enough to spot and didn't secretly propagate themselves through thousands of lines of code.

Made a pr to change a thing after i told it specifically where and in which repo. It made the pr in wrong repo.

Suggesting that, since bleach is a great cleaner, and ammonia is a great cleaner, mixing the two would make an even better cleaner.

Work demanded we lean heavy into AI, bought into ChatGPT early

Our first townhall meeting after that was so cringe. They made a bunch of uncanny valley slides up with amazing numbers how they were ruling the industry, They even included pages of citations.

Every single citation was a 404. Half the numbers given for the rest of the industry were decimated versions of their real values.

They spent a lot of money on tokens to unknowingly make up a deck of lies, then presented them to thousands of people.

Last year, 2025, Google's Ai overview confidently stated that four pieces of 26 gauge wire twisted together would be the equivalent of one 32 gauge wire. For those that are unfamiliar with wire gauges, a higher gauge number means a smaller wire.

Not personal but the instances where AI tells people who are suicidal to go through with it or recommend people do things that ultimately kill them.

Not exactly life-critical but on games I play I've asked for optimization advice and it has responded with info I know is 100% wrong. I end up replying with URLs to the info and pasting the direct info from it. The AI of course goes, "Of course!".

At work I constantly get commands that dont work in my version of some app or OS even after I told it the exact version I'm using. I'll even post the response from tech support and it'll throw up its virtual hands and say "of course!".

I've also noticed that AI has now been around long enough to start posting clearly outdated information.

I'm forced to use it at work and my token usage is monitored. Here was yesterday:

Build things thing, but remember: DO NOT DO THE THING.

Sure thing! I'll build out and I will not do the thing

WTF YOU DID THE THING

You're absolutely right, I did the thing even though you explicitly told me not to do the thing

I wonder if they will have metrics for how many times I've all caps sworn at it.

I haven't really had a good experience with AI yet.

Well, it seems that every time a search engine's AI gives me a definitive and wordy "answer" to a question I search and provides a helpful citation of the page from which it got that answer, it is ALWAYS just a completely hallucinated answer that the cited page in no way supports. This isn't just hyperbole, this has been my experience with every single search engine AI summary

And I find myself having to go back and fix the autocorrect that my phone keyboard "helpfully" makes with increasing frequency. It used to be that autocorrect would just fix a typo when I'd obviously typed gibberish because I missed a key. But now, it is regularly replacing entire words that I've typed completely correctly with some other word that makes absolutely no sense.

I'm all for some specific uses of AI. Things like speech to text or images to text? Sure that's a great use. But when it comes to research, writing, or spreadsheets, AI tools just need to shut up and stay the fuck out of my way. I'm becoming convinced a little more everyday that anyone using AI to summarize articles or search results is a complete moron and shouldn't be trusted.

I kinda disagree with that, AI is rarely doing mistakes for me, atleast Laguna S2.1 and Mistral Small 4