Bold statement, huh?
Sometimes we go a little 'overboard' with the new tech tools, don't you?
Like our friend AI. We LOVE AI. Really, we do. There are so many great things about AI. We know about all these amazing benefits; people are talking them up all the time.
Faster, smarter, better!
However, for all the 'wins' we see, the fact is, AI has also created an interesting little problem.
Now that we can use AI for almost everything, we're starting to assume we should.
Think about ALL the things you do daily: organization, automation, information. Everyone wants to shoehorn AI into these tasks, even if it's not the right fit. Or, more accurately, even if it isn't needed.
It's actually ok to be less exciting.
You heard me. Be boring. Go simple rather than flashy.
Don't let AI complicate things that you already have a solution for......... you might not be using it, but you do have a solution for it.
Start with the problem, not the solution
You might be asking the wrong question entirely. Before you ask, "How could we use AI here?", think about it a little differently.
Try asking a different question. Like, "What problem are we actually trying to solve?"
Maybe someone needs to find information faster.
Maybe a repetitive task takes 20 minutes every Friday.
Maybe employees keep writing the same email.
Maybe information needs to move from one place to another.
Once you understand the problem, you can choose the right level of technology.
Remember the AI Escalation Ladder?
We introduced this a couple of blogs ago. When it comes to AI, I'm taking you from the bottom up. The solution might be so much simpler than adding more tech.
Start at the bottom, and only climb when the next level is a value-add.
Here's a few examples:
Level 1: Existing functionality
Could an Outlook Rule, Excel feature, Teams feature, or another capability you already own solve the problem?
If yes, stop.
Level 2: Copilot assistance
Would a Copilot prompt help someone complete the task faster?
If yes, you might already be done.
Level 3: Repeatable automation
Does the process happen frequently and predictably?
Now something like Power Automate may make sense.
Level 4: Reusable AI capability
Do multiple people repeatedly need the same AI-supported process?
Now we can think about reusable prompts, shared solutions, or more sophisticated AI workflows.
Level 5: Agents
Does the situation genuinely require greater autonomy, multiple steps, system interaction, or ongoing execution?
Now we're having an agent conversation.
The goal isn't to reach the top of the ladder, it is to SOLVE your problem.
I love games, and I love to win. But that's not what this is about. You do not get bonus points when you reach the top of the AI Escalation Ladder.
You win when you use the tools you already have and actually solve your problem!
Imagine that.
If an Excel feature solves the problem, that's a win.
If a good prompt solves the problem, that's a win.
If a simple automation works, stop there.
The goal isn't maximum AI.
The goal is maximum usefulness with the minimum necessary complexity.
This matters for you and your organization because....
.... when you layer on more complexity, you get more maintenance. Training needs. Dependencies. Governance.
And risk. You get more risk.
That doesn't mean that you never need that level of complexity. It is there for a reason, and it can be absolutely worthwhile.
It can also be completely unnecessary.
The smartest AI users know the difference. This is becoming an AI skill of its own.
I want you to embrace AI. I want you to be thoughtful about it and don't fall for the shiny-tool-syndrome and confuse impressive with effective.
The users that will find the most value are the practical ones.
Not flashy. Save that for another day.