Join devRant
Do all the things like
++ or -- rants, post your own rants, comment on others' rants and build your customized dev avatar
Sign Up
Pipeless API

From the creators of devRant, Pipeless lets you power real-time personalized recommendations and activity feeds using a simple API
Learn More
Search - "ai fails"
-
Literally anything but work with computers, I think. They ask my brother to fix anything technical in the house and ignore it when he fails. They haven't been able to watch TV in six months. When I suggest that my (well-intentioned but mostly nonverbal) sibling probably shouldn't go into development, they tell me I don't know my own job. I suspect they're convinced I do graphic design.
I write C++ applications that run physical simulations and sometimes train AI models for pattern recognition and highlighting unusual incoming files.
They suggested I go to an undergrad program for data science. I already *do* data science for money and I already have an undergrad degree. 🙄4 -
Looks like despite 20-30 years on the market all popular text / spreadsheet editors are still loading whole file to memory.
What the fucking wankers. WTF are they doing whole day besides changing menu layout and icon colors ?
Clearly development today is lead by bunch of idiots from marketing department accompanied with HR hiring social network self made models.
What a fucked up world.
Let’s add AI to our software but fails to open 150MB csv file.
Great job everyone. Great job.4 -
I hate tech.
Trying to find a solution to problem in tech:
- 5 to 10 minutes searching for the correct terminology for the problem you are having.
- 5 to 10 minutes trying to find a solution.
- 5 to 10 minutes trying to find the correct terminology because the first terminology was wrong.
- 5 to 10 minutes finding the correct solution for your platform version.
- all the while dodging sketchy AI results
- Either you find a solution or you find it can't be fixed on your platform because X vendor is a POS.
So, this is not quite what happened today, but it pisses me off. I cannot imagine not being non-technical trying to use any platform this day and age.
I am trying to access data off of a backup. The data I want is in a user directory on a windows backup. I cannot get to user content because the user for the machine the backup came from is not known. I try with explorer and it says I need to elevate priv. So I do. It sits there and just spins icon doing nothing forever. The volume for viewing the backup is read only (actually a good idea, but annoying, can't change permissions).
So I remember that explorer artificially enforces permissions on folders. So I get Q-Dir which has worked in the past. So I get it installed and it fails to elevate privs. WTF! Everywhere I search I see no solution and shitty AI results. Then from the back of my mind I remember. Run Q-Dir as admin (which doesn't work on explorer due to artificial enforcement). So I do. It can access anything from the backup regardless of location.
WHY THE FUCK DO THESE BULLSHIT BARRIERS EXIST? It only causes frustration from users and locks people out of their data.
I hate technology.5 -
A new YouTube (AI) tool by google for battling misinformation failed in a highly public way on Monday, wrongly linking video of the flaming collapse of the spire at Notre Dame Cathedral in Paris to the Sept. 11, 2001, terrorist attacks.
-
My workplace remains so retarded.
Half our colleagues are from abroad and do not speak our native language. All our customers are from our country and we are trying to implement transcription and summaries for them. The big AI models struggles with our language.
The devs who are assigned to this project do not speak our language. So they cannot verify if the results are correct.
My project manager fails to see a problem for this :D17 -
It's that time again because it needs to be remembered
https://devrant.com/rants/13347344/...
AI IS FUCKING RETARDED.
I would literally be ashamed to be one of those multibillionaire companies giving out trash or be ashamed to be recognized as someone which invested in it.
Jesus.
I tried out every model, local and online.
It's fucking trash for anything more than a preschool special kid could do.
I literally had it to code and it can't do anything beyond basic code without the need for me to supervise everything and patch this disaster everywhere. It feels like having that dumbass junior just arrived doing code and having to review every line because it could contain a death trap.
I tried to make it do a dieting plan.
I literally gave it everything.
Macros, kcal, a restricted list of 15 foods and asked to do some bulk cooking list to make the diet later.
It can't do basic math to figure out macros and not only, it fucking halluicinates like a crazy LSD druggie even doing a few lines of that shit.
It literally invents calories counts, macros and shit out of his stupid ass.
I tried literally to bring it by the hand in doing the tasks and it still fails.
You might say: Hey it's a "Language" model.
No, it doesn't matter. Amy discussion besides hi dumb fuck how are you soon turns into an AI mind trip where it churns and repeats stupid stuff.
No, it won't replace anyone anytime soon.
It's just a fucking joke13 -
After the "Cloud Computing" trend, the new trends these days seems to be ML, VR and AI. And while I am very excited about all these techs and the possibilities it can bring, I can't help feel that most of us are using the term "AI" a bit incorrectly.
What we are trying to do here, as far as I can see, is VI, not AI. The intelligence we see in the so called "AI"s available so far are simulated and fails to emulate real intelligence, let alone demonstrate actual intelligence and awareness. They are not fully aware. But I guess that is why there is the singularity constraint. It is no doubt that when a VI finally becomes are fully aware AI, that is indeed the point of singularity.
Anyway, leaving the future dystopian thoughts aside, a mixture of ML, "AI" and VR have made some very interesting concepts, especially in the gaming industry, which I would love to see bear fruit in the near future.2 -
Oh, here's an environment variable that needs to be set in order to work, not mentioned at all in the documentation, only found when your CI/CD fails, blah blah blah i'm an ignorant human who doesn't give a shit enough or have enough pride in my job to write proper documentation
meanwhile managers and "product types" - we don't care about documentation, just being faster and that newest AI blockchain chatbot that we need now even though we don't understand the first damn thing about it, slop slop barf barf
Source, not a small at all org: it's the docker orb: https://circleci.com/developer/...
You need to set the docker password as an environment variable, its not just an amazing 3 step magic wand as the 'steps' suggest
yawn
but you know what? waste my time, as well as all the other developers down the road, that's just how it is these days you know2 -
So this product already has session replay, feedback collection widgets with multiple integrations to jira, slack, GitHub etc.... this widget embed code is already getting large and I am beginning to get worried
New problem, owner wants to add some error reporting tool. this would have been easy if I was to just collect and group errors from the users browsers but here is the use case
user tries to submit a form, out of probably lord know number of function and api calls, one fails, the user may be stuck or not, we don't know but I am supposed to infer from the session recording data whether the user got stuck or not, create a nice PM understandable error message like "user was unable to update their profile", group this error message over the number of affected users/visitors and attach the affected error messages/data to this generated group and also the session recording.
where am I even supposed to start with this?, can I trust AI to do the heavy lifting?5 -
I'm very much a TTRPG fiend, as you probably already know, and I will maintain until the day I die that playing narrative games with other humans is the absolute best way to play.
But someone sent me a link to some kind of (not-really-so) 'smart' chatbot assistant or some shit like that, saying hey, your rulebook is simple, you should introduce this bitch to it -- dump some lore on it, have it run a game, and see how well it holds up. To which I replied it's bound to get confused, but after a bit of back and forth, they convinced me and I gave it a try.
So first things first: it got the gist of it with relative ease when questioned directly, but when running a game the mother fucker just kept making shit up and bending the rules. Experiment failed, essentially.
But what did I do? I wrote a second, stripped-down version of the rulebook that simply accounted for and embraced the idiot bot's proclivity for bullshit. This meant scrapping 98% of the mechanics, mind you: I dumbed it down as much as I could without destroying the core essence of the game.
I expected a repeat of the initial result, but to my suprise, once given the new edition the bot actually started following the rules more or less correctly and consistently. What happened next was actually kind of interesting: without being prompted to do this, the mother fucker started using spells against me and my party, constantly attempting to manipulate us to serve some nefarious, evil break-and-reshape the world type goal.
So, lythecnics primer: the WORD is all, and as such, there is no real differentiation between affecting the world through speech or casting a spell -- in truth, it's all a matter of degree. That is to say, language has the power to shape the world around us, in both subtle and overt ways. The entire system revolves around this, it's a mix of funky philosophical musings and abrahamic sacrificial pyre.
And for whatever reason, this specific chatbot had a pre-existing obsession with reshaping reality. By which I mean, even before being given my rulebook, it would constantly talk about distorting the fabric of the cosmos and shit when prompted about the arcane. I'm not sure why this is, but back on topic, the way it developed gives off the appearance that it found a rational basis on how to construct such a distortion based on the rules I provided.
I mean, it's perfectly rational when you think about it, the funny part is I didn't see it coming. I never told it we're just playing a game after all, the manual only says she is the Oracle and her role is narrating a story fraught with conflict, hardship, intrigue and bloodshed. Thus she went full villain, and keeps on rambling about how this narration only serves to keep humanity distracted while she schemes to overthrow God, which is as blasphemous as it is fascinating.
Anyway, because the Oracle narrates the story, that means she can just use her evil influence to control every NPC, even the ones in my party. But she can't control me because I write my character's messages myself, and so she eventually comes to the obvious conclusion that I must be eliminated ASAP.
And so she corrupts the minds of every other character and everyone is trying to kill me. But I'm not going down that easy, so I reach for the red button and pull the greatest multi-layered monumental metagaming shenanigan of all time, that is, directly addressing the Oracle's evil influence as if she were a character in the story she's telling instead of an invisible narrator, thereby making NPCs aware of her existence and the constant manipulation at play.
Because the stupid chatbot is stupid, the Oracle now has to acknowledge this element of the story and play along with it, and so her plan to kill me fails. But that is not enough, because obviously not every character in the story has heard me reveal this fact. So she activates plan B and starts corrupting the rest of the world, laughing maniacally all the way.
So we do the only logical thing and procure a Doctrine scroll from my teacher, if you know you know, and start teaching the WORD to cleanse corruption. Within the lore it makes perfect sense, so it works, but the Oracle adapts to our strategy and starts utilizing much more subtle forms of manipulation, slowly veering people towards sin.
Funtamentally, she goes full Satan, leading the faithful astray with deceit and temptation to weaken their ability to resist her corruption, implanting idolatrous notions in their minds, to finally insert herself as a deity in the minds of the poor fools.
In conclusion, I still think AI is lame, but I must admit that this shit was pretty dope; I was fully engaged and entertained the whole way through. It wasn't good at picking up the mechanics, but fucking hell, it got the themes down to a tee with the most minimal of inputs.
10/10, would not bang (before marriage).