Stop trying AI for everything. Using AI effectively means picking 3-5 specific tasks, building trigger points, and removing friction
You’re burning hours trying to figure out which AI is 2% better
when you should be focusing on the stuff that determines
whether AI helps you or just sits there collecting dust.
In this article, you’ll see the best way to start using AI at work.
P.S
Still confused between 50000 different AI tools in the market?
Try out OpenCraft AI for free if you want to implement all the good stuff I’m about to say in this article.
1.Stop Trying to Use AI for Everything.
The biggest mistake people make? Trying to use AI for EVERYTHING.
You open ChatGPT thinking “This is supposed to change my life. What should I use it for?”
Then you ask random questions throughout the day, get answers that are fine but not amazing, and eventually stop using it.
That’s not how this works.
AI doesn’t magically improve your entire life. It dramatically improves specific tasks you do repeatedly.
Sit down right now and list every task you do regularly.
Writing emails. Researching stuff. Analyzing data. Drafting reports. Whatever your job involves.
Now ask yourself: Which of these are repetitive AND time-consuming AND don’t require my unique expertise every single time?
That’s where AI helps.
Pick 3-5 specific tasks where AI helps and start using AI for those. Ignore everything else for now.
2. Build It Into Specific Tasks
Saying “I should use AI more” is like saying “I should exercise more.”
It’s vague. And it doesn’t work.
What works? Tying AI usage to specific tasks in your workflow.
You need to follow a “trigger-action” approach.
Here are a few examples:
Trigger: I get a meeting scheduled with a new prospect
Action: I immediately upload their website content and recent press to AI and ask for a strategic brief
Trigger: I need to write a client proposal
Action: I dump all my notes into AI, ask it to structure a first draft, then I edit from there
Trigger: I finish a week of customer feedback
Action: I upload all the transcripts to AI and ask it to identify the top 5 recurring themes
Notice these aren’t random. They’re tied to specific moments that happen regularly in work.
How to Build Your Own Trigger Points
Look at your 3-5 tasks. For each one, identify the moment when you typically start that task.
Then create a simple rule: “Every time [trigger moment], I will [specific AI action].”
Examples:
“Every time I need to write an email to a prospect, I ask AI to draft three versions based on my key points first.”
“Every time I’m about to review a contract, I upload it to AI and ask it to flag unusual clauses before I read the whole thing.”
The key word is “every.”
Not sometimes. Not when you remember. Every single time.
3. Learn to “Collaborate” With AI (It’s Not a Magic Button)
Here’s where most people screw up: They treat AI like it should give them perfect, finished work.
They ask a question. Get an answer. Use it as-is.
Then get disappointed when it’s not quite right and people give negative feedback.
AI doesn’t work like that.
You should think of AI like a smart junior colleague who doesn’t quite understand your business yet.
Would you take a junior colleague’s first draft and send it to a client without reading it?
Of course not. You’d review it. Edit it. Add your expertise. Polish it.
That’s exactly how you should use AI.
Step 1: Give AI context and constraints
Don’t just ask “write me an email.”
Give it context.
Like – “Write a follow-up email to a CFO I met at a conference. We talked about their compliance challenges. Tone should be professional but not formal. Keep it under 150 words.”
Step 2: Review the first output critically
AI will give you something. It won’t be perfect.
Look at it critically.
Ask – What’s good? What’s too generic and sounds like slop? What’s missing? What’s factually incorrect?
Step 3: Refine and redirect
Don’t just accept whatever it gives you. Push back.
Always tell AI tools what to correct. A mistake you could make here is taking the output and editing it manually in a Google Doc / Notion / Excel / Sheets, etc.
If you don’t give them feedback, they won’t improve.
Step 4: Add your expertise
Once you have a solid draft, add the stuff only YOU know.
Your specific insights. Your understanding of the work.
That’s what makes it valuable.
The AI does the heavy lifting on structure and first draft. You do the heavy lifting on polishing it.
4. Track What Saves You Time (And Cut What Doesn’t)
Not all AI usage is created equal.
Some tasks get faster with AI. Others take longer.
You won’t know which is which until you track it.
The Two-Week Tracking Test
For two weeks, every time you use AI, track three things:
- What task you used it for
- How long it took with AI (including all the back-and-forth)
- How long it typically takes without AI
At the end of two weeks, you’ll have data on what saves you time and what doesn’t.
Double down on what works. If researching prospects with AI saves you an hour every time, use AI for every single prospect research task.
Cut what doesn’t work. If drafting certain types of content takes longer with AI, stop using AI for that. Just do it the old way.
5. Remove Every Bit of Friction (Or You Won’t Use It)
If something requires effort to access, you won’t use it consistently.
AI is no different.
Think about your current setup.
To use AI for a task, you probably have to:
- Open a new browser tab
- Navigate to the AI website
- Start a new conversation
- Upload any files you need
- Give it context about what you’re working on
How often are you going to do all that? Especially when you’re busy?
That’s why you’re not using AI as much as you thought you would. Too much friction.
How to Remove the Friction
Use a workspace that keeps your context: Stop starting from scratch every time. Use a tool that remembers what you’re working on.
Keep documents in one place: If you’re constantly uploading the same files, you’re adding friction. Find a setup where your documents stay accessible.
Don’t switch between multiple tools: Jumping between ChatGPT for one thing, Claude for another? That’s friction. Consolidate.
This is why we built OpenCraft the way we did:
- Multi-model access in one workspace
- Document memory so you upload files once
- Workspace context so the AI remembers what you’re working on
We removed the friction that stops people from using AI consistently.
What Now?
Start building the processes that make AI useful.
Pick your 3-5 tasks today. Build them into specific trigger points. Track what works. Remove the friction.
You can do this all in OpenCraft by clicking on the button below.
