Skip to main content

People think they are helping by showing me what their AI Chatbot said, but it just doesn't...

· 3 min read

Ever feel like you’re the IT version of a personal trainer, but everyone’s just showing you the workout plan from a fitness app instead of actually telling you what they want? That’s the daily grind for our hero here, who’s been receiving a steady stream of “I just asked ChatGPT and it said X” from end‑users. The irony? The answers are often as stale as last week’s pizza, or worse, they’re from a model that was last updated before the coffee machine stopped working.

The “AI‑Assisted” Problem Solving Cycle​

  1. User: “I need an incoming webhook on Teams. I asked ChatGPT, it gave me a solution. Tried it, still no good.”
  2. IT: “Let’s hop on a call to figure out exactly what you tried. Maybe the bot didn’t know about our firewall.”
  3. User: “I need a new laptop for Photoshop. The LLM listed the latest model, but it’s discontinued.”
  4. IT: “Okay, I’ll do a quick search and get it approved. Yes, because the AI can’t just know inventory levels.”

The result? A half‑finished support ticket, a half‑satisfied customer, and an IT professional who’s basically doing “AI‑assisted” troubleshooting while feeling like a time‑traveling librarian.

Why the AI Is Not the Answer​

  • Outdated info: The model’s knowledge cutoff is 2023. A “new” laptop from last year is now vintage.
  • Context‑less advice: The bot can’t see your network’s quirks, your team’s policies, or the fact that your Teams channel is protected by a corporate firewall.
  • No empathy: It can’t read the frustration in a user’s voice or the subtle hint that the user actually just wants a quick fix.

In short, the AI is like a brilliant but clueless tour guide who keeps pointing at a monument that’s been replaced by a parking garage.

Community “Cheezburger” Reactions​

But… but… AI! /s
Also, these are the people who ask ChatGPT instead of their doctor as well. At least soon enough they won’t be asking anything anymore.

Or (even more frightening for society) asking AI: What should I vote?
Asking an A.I what you should vote is just called Managed Democracy(tm).

Back in my day it was always “My mate down the pub says…”, but never as bad as this.
I think part of the problem is that corporations are heavily pushing their use in regular workload. So since it becomes a regular approved tool then they use it for everything and figure that’s a good solution.

We’re all just waiting for the day when the AI gives us a life‑changing career plan and a budget spreadsheet that actually matches our real‑world constraints.
Until then, keep those “I asked ChatGPT” replies coming. They’re the real comedy gold.

TL;DR​

IT pros are getting “help” from ChatGPT, but the AI’s advice is usually outdated, out of context, or just plain wrong. Users keep sending the AI’s output, IT keeps correcting them, and everyone ends up with a story for the next office joke. Meanwhile, the AI remains blissfully unaware that it’s not actually solving the problem—just adding a new layer of confusion.