
- My Forums
- Tiger Rant
- LSU Score Board
- LSU Recruiting
- SEC Rant
- SEC Score Board
- Saints Talk
- Pelicans Talk
- More Sports Board
- Fantasy Sports
- Golf Board
- Soccer Board
- O-T Lounge
- Tech Board
- Home/Garden Board
- Outdoor Board
- Health/Fitness Board
- Movie/TV Board
- Book Board
- Music Board
- Political Talk
- Money Talk
- Fark Board
- Gaming Board
- Travel Board
- Food/Drink Board
- Ticket Exchange
- TD Help Board
Customize My Forums- View All Forums
- Show Left Links
- Topic Sort Options
- Trending Topics
- Recent Topics
- Active Topics
Started By
Message
re: AI garbage in garage out
Posted on 8/1/26 at 9:37 am to Dallaswho
Posted on 8/1/26 at 9:37 am to Dallaswho
quote:
That’s what dependency bot is for.
Yeah solutions like that have existed for a long time... yet here we are.
I think it is being understated how much of the underlying stuff is being updated with massive feature sets that are growing exponentially larger.
I dont disagree with the OP, that minor stuff like this probably is low risk.
But the dependency tree is getting deeper and wider, with code bases that are being shipped with little to no testing.
Posted on 8/1/26 at 2:34 pm to jdd48
quote:
Great way to kill his confidence and teach him the wrong way to go about things IMO. He's actively trying to learn as a junior and you're telling him "nahhh... just use the magic answer box."
Junior to me, not inexperienced. The guy has been with the company over 10 years. Despite all that he insists on not changing with the times.
Posted on 8/2/26 at 12:16 am to jdd48
quote:
Critical thinking and overall deep process understanding will be lost at some point if everything just comes out of the magic answer box.
I am floored at how much AI slop is coming out of people in the workforce now; it's impressive at first, but after a while you start to notice it: the sing-songy this not that style in the prose, the too small text in the documents, the pointlessly complex code with overtly verbose code comments.
I've also noticed some problems with the magic answer box as of late, especially if it's something not easily found on the internet, even if you do provide it context. It's still helpful, but it's not the gospel it's portrayed as, and the monthly flat rate has been changed to a meter in a lot of places.
I keep getting told my job is going away by the media; never say never, but it won't be because of AI.
Posted on 8/2/26 at 8:59 pm to GetMeOutOfHere
This post was edited on 8/2/26 at 9:07 pm
Posted on 8/2/26 at 11:48 pm to CAD703X
I don't even know what this green check mark means
Posted on 8/3/26 at 7:44 am to HailToTheChiz
quote:its a lazy-arse way to generate a to do list for other people or make it seem like you actually have a handle on the current project when all you did was ask AI to summarize everything.
I don't even know what this green check mark means
Posted on 8/3/26 at 2:33 pm to hob
quote:
Junior to me, not inexperienced. The guy has been with the company over 10 years.
Yea - don't do that then. He was sure flipping you off through the monitor.
Posted on 8/3/26 at 3:00 pm to Dallaswho
quote:
That’s what dependency bot is for.
GHAS remediation is exactly what I'm doing right now. Since the overwhelming majority of vulnerabilities reported are Dependabot, I asked if we could auto-update so the repo owners could concentrate on code scanning and secrets.
I was told that the auto updates had been turned off at the request of the coding teams that figured out the dependencies were being auto updated. It caused no problems, and none of them even noticed for six months. But turn it off, they did.
And one of the worst organizations with repos with security problems is Company-AI-Engineering. Claude was responsible for nearly all of the code problems in the top 10 repos in that org.
Posted on 8/6/26 at 9:22 am to LSURep864
quote:
Amazing yet Stupid
Indeed free-level Copilot and Gemini amaze. One found a PDF of an esoteric Russian publication, and answered my questions about the content. I do worry about how many answers were actually true. I asked another to remind me how to execute a task in Excel, and it replied that I should upload the file and it would do it...which it did.
The stupidest thing is that both agents are usually wrong when it comes to customizing either Windows11 or Chrome OS. I'm beginning to think Microsoft and Google enjoy confusing their users.
Posted on 8/6/26 at 2:41 pm to Tree_Fall
quote:get a paid subscription to cursor or claude or chatgpt.
The stupidest thing is that both agents are usually wrong when it comes to customizing either Windows11 or Chrome OS. I'm beginning to think Microsoft and Google enjoy confusing their users.
chatgpt searched for the perfect deal on my hardware, got me a KILLER deal on ebay, diagnosed and checked all equipment out, helped me wipe windows 11 and install proxmox, a lightning fast windows vm (seriously..it boots from cold start to login in 8-9 seconds), tailscale and now HA and plex installs.
i dont think its screwed up more than 2 or 3 times with the tasks. i'm not doing 1% of what AI is capable of here but even things like the powershell script it gave me to install the core apps and utilies i wanted was so freaking sweet.
i didnt realize how annoying web searching and finding the 'download' button link was until i had a powershell script that installed 8 different tools from all over the web and all i had to do was hit "ENTER".
Posted on 8/6/26 at 8:31 pm to hob
quote:
Subject matter experts that know how to write good prompts are going to rule the world.
You realize that AI can be subject matter experts and write good prompts too, right?
Posted on 8/6/26 at 9:32 pm to tigrewoulds21
quote:
You realize that AI can be subject matter experts and write good prompts too, right?
Yes, it can be, but it can also be full of shite. No Skynet yet.
Posted on 8/7/26 at 7:54 am to CAD703X
Great tip. My prompts should be asking for scripts that fix my problems rather than asking for how I can fix them.

Popular
Back to top


0










