Everyday AI for everyday people. Not those of us who are paid to flirt with new tech. The other people. Those who often have it thrown at them, or see a Super Bowl commercial a use a phone-a-friend to find out more.
I’ve spent the last little while watching people discover AI in their own slightly chaotic ways. You’d think, given all the hype, we’d have a grand plan. Instead, most tiptoe in like they’re poking a suspicious container in the back of the fridge.
And almost every time, the first thing people do is ask AI to tell them something.
Not surprising. We spent two decades learning to type questions into search engines and hoping the answer we need is somewhere between the sponsored ads and the recipe blog with the life story no one asked for.
So people will ask AI things like:
“What’s inflation?”
“Where should I eat near Paddington Station?”
“Is my dog broken or do all dogs hate ceiling fans?”
Every one of those is fine. Normal. Workable. Expected.
But it’s only part of the story.
Then you accidentally stumble into “do this for me”. It might happen by mistake. Maybe you’re tired. Maybe they hit a deadline where “figuring it out yourself” stops being noble and starts being impractical.
And they say something like:
“Can you rewrite this email so I don’t sound like I’m begging for mercy?”
or
“Here’s a wall of notes. Make it make sense.”
or
“Turn this into a PowerPoint because I cannot stare at one more blank slide.”
(Side note: these prompts aren’t that bad. Do this for this reason. Solid prompting)
And suddenly AI isn’t a search engine anymore.
It’s a worker.
A partner.
Someone you hand a messy idea to and say, “please just… start.”
I remember prepping a conference session recently — the kind where I’m trying to be helpful and insightful but also entertaining enough that no one wanders off for more coffee. I had too many ideas, no shape to most of them, and certainly not enough time. So, I opened all my AI tools like I was loading up a spaceship cockpit.
First I asked for information.
“What’s a simple way to explain this concept without sounding like a textbook, or looking like an idiot?”
“Is this analogy terrible or just mildly confusing?”
Then I handed over the messy parts and said, “Here. Fix this.”
“Rewrite this section as a story.”
“Turn this outline into something that looks like I meant to write it this way.”
“Make an image that’s actually usable and not emotionally disturbing.”
None of that would have worked if I’d stayed in the “tell me things” lane.
And none of it would have worked if I’d skipped straight to “do things for me” without understanding where I was heading.
For those who ask for help, it gives direction.
For those who ask for tasks, it gives momentum.
You need both.
Somewhere along the way we convinced ourselves AI is a genius that should be treated with reverence, when in reality it’s more like a very fast, very caffeinated intern. Sometimes it gives you brilliance. Sometimes it gives you chaos. But it always works better when you give it something to do with your starting point.
I watch this play out in regular life, not just tech circles. Someone planning a trip will ask, “What should I do in London?” but maybe not think to say, “Create a walking itinerary starting at Paddington, with food stops and a downloadable map.” Someone looking for a meal plan will ask what toddlers can eat instead of saying, “Make me a vegan meal plan for two adults and a toddler that I can prep on Saturday.” Someone reading a never-ending email chain won’t say, “Summarize this in five bullets and tell me what actually matters.”
(Ok this was my first genuine foray into using consumer AI tools. I have a friend who writes super long, well thought out emails. And I don’t need ALL of it, just the highlights. “Summarize this email” was my go-to prompt for a while there. Sorry Shan, you had a lot to say before you retired!!)
It’s not that people don’t want help.
They just maybe don’t realize they’re allowed to ask this way.
And here’s the part I wish more folks understood: you don’t have to choose the “right tool.” Copilot, ChatGPT, Midjourney, Gemini, Claude — they all have their strengths, and the secret is that you can mix them together like ingredients rather than pledging allegiance to one platform like you’re joining a medieval guild. I use ChatGPT all-the-time to take my crazy ideas and craft a prompt to give to Copilot to help me build stuff in Power Platform.
Start with an idea in one place.
Shape it somewhere else.
Pull it into PowerPoint.
Touch it up.
Rewrite it.
Hand it back.
Let it pass through tools the way work passes through people.
Most of us are already doing this in other parts of our lives. AI isn’t special — it’s just new enough that people forget they can be a little bossy with it.
AI needs the human element to be successful. And that’s what you are doing.
If you want to help people level up, the easiest place to start is teaching them the magic phrase:
“Here’s what I have. Do something with it.”
It’s not about crafting perfect prompts or hitting the right tone. It’s about letting both camps — the “tell me more” camp and the “do this for me” camp — meet in the middle.
That’s where the real usefulness is.
That’s where people shift from dabbling to actually getting things done.
And once you see that?
Once you feel the difference between information and transformation?
There’s no going back.
Here’s the conversation I had with my ChatGPT to help craft this blog post. I use it this way often. Here’s my crazy idea, let’s make it better…make it organized…is this a viable something… A few things to notice:
- Months ago, I asked ChatGPT what its name was. It replied “Gio.” So, make sure you find that part where it refers to itself in the third person offering to do something for me.
- It’s eager to DO something. It only asks questions when told to ask questions.
- We iterate. It goes fast; I slow it down. Sometimes it just takes off, when I need to toss around a few more ideas first. But it’s coachable.
- It likes lists.
- It asks way more questions than it should. But usually, they are good questions.
- I’ll give it the finished text for reference later, so it knows what I did with our draft.
Leave a Reply