What’s the deal with AI and Automation?
They’re not magic potions.
We recently had the opportunity to share examples of using artificial intelligence (AI) in our content workflows via the Community of Practice session with communications officers from the Ministry of Home Affairs.
AI is fair game for everyone. Although we may be at different stages in how we use it, there’s always something to learn from one another.
The other day, we tried moving a “LinkedIn-to-Viva Engage” Power Automate workflow between user environments. I told the engineer I wanted to use JSON info from an email to automatically update a SharePoint Excel as we have been doing that manually.
I saw his eyes lit up. “I can help you with that,” he said.
Turns out, the solution already exists. I was overcomplicating things.
Combining AI and automation can help reduce manual and repetitive work.
In short, AI “thinks” and automation acts.
It’s so easy It’s not. It takes multiple tries to get the results you want. For example, I often have a vague idea for an image but don’t know how to write the correct prompt. So, I asked ChatGPT to write it for me. While automation can save time, it takes considerate effort to build clear if-this-then-that logic. Poorly designed workflows can cause more problems than they solve.
It’s not a solution looking for a problem The challenge is setting aside the time to identify what you do repeatedly every day and break down the steps.
AI will replace jobs and humans In reality, it’s more likely to change jobs than eliminate them. Most AI today still needs human oversight. For example, some tools let you generate a video using text prompts. In programming, some developers call it “vibe coding”. However, if you have a working knowledge of video editing software, you’ll get better results. You’re not replacing the skill but using AI to make it more efficient.
AI is objective Not really. AI is trained on human data, which can be biased. If the data is flawed, the AI will be too.
More data does not mean better AI It helps, but the data must be relevant and high quality. Too much noisy or biased data can make AI worse.
As for ChatGPT, these are custom traits I told ChatGPT it should have:
Use simple language: Write plainly with short sentences.
Avoid AI-giveaway phrases: Don't use clichés like "dive into," "unleash your potential," etc.
Be direct and concise: Get to the point and remove unnecessary words.
Maintain a natural tone: Write as you usually speak.
Avoid marketing language: Don't use hype or promotional words.
Keep it real: Be honest; don't force friendliness.
Stay away from fluff: Avoid unnecessary adjectives and adverbs.
Focus on clarity: Make your message easy to understand.


















