How to Get Better At Delegating to Your AI Assistant
Thanks Dorothy! Look, a ton of us, are all learning how to work the best with our AI assistants. I know I am. If you are too, this one is for you. Whatever your job may be, or even how seasoned you are at your role, yo
Thanks Dorothy!
Look, a ton of us, are all learning how to work the best with our AI assistants. I know I am. If you are too, this one is for you.
Whatever your job may be, or even how seasoned you are at your role, you still might be new to delegation.
Well let me try that again, youβve definitely been delegated to, but you might not have ever had the chance to delegate to someone else. Thatβs totally normal, it can feel foreign. Even managers need to re-educate themselves often.
Itβs a skill, and you will get better at, it the more you practice it.
Before we practice though, letβs break it down.
A 90βs inspired motivational poster for you I imagenβd. Motivated?
Delegation means entrusting someone with work, giving them enough context and permission to move that work forward, and agreeing upon when to check back in on how the work went.
The best way to practice delegation is to do it. Start with something you know how to do but keep putting off.
Universally, one thing I am fairly certain every reader here could benefit from is building a tiny week in wins report for themselves. A brag journal if you will. We all do a lot of work and sometimes we forget to celebrate our wins. Not only do we know how it is important for our own self growth, we also all realize that it's a wonderful way to stay motivated to accomplish our goals. But alas, we donβt do it. (If you have a weekly wins report already, find a new thing along these linesβ¦ but also, Iβm super impressed, great job at life!)
And because we donβt look at these wins often enough, Iβve personally found, things Iβve done, that I spent my time on, whether that be creating or even learning something new, fall through the cracks. I forget about all the things I do in a week.
So let me show you how I do this now with my personal assistant.
Dorothy, Dot for short
This is Dorothy, my personal assistant, my dot. During onboarding we got to know each other pretty well, so I gave her some context, about what Iβm up to, what I care about, yβknow things like that. Most assistants have a similar onboarding experience.
Make use of connections from the apps where you work
With Dots, a great way to get things going is to make use of connections. You connect your Slack conversations, meeting notes, Gmail, Calendar, and Drive. You can point your assistant towards the relevant projects and sources. If something isnβt yet connected, you can always just share a few notes or links to give it a peek into what you are up to. Dorothy even has her own computer and can connect to and use mine directly.
Start with a simple assignment:
βHelp me prepare my week in wins. Show what I accomplished, including things I learned or experimented with, who helped, and why it mattered. Use the sources we agreed on, ask about anything missing, and keep it private until I review it.β
You can add the details as you go. Whatβs finished and whatβs still in progress? Why did this particular item matter? Is there evidence for the impact, or is that something youβre hoping to learn?
Then look at what comes back.
βYou missed this project.β
βGive my teammate credit for that.β
βThis looks small, but hereβs why it mattered.β
That conversation is part of the work. Youβre practicing how to explain your priorities, give useful feedback, and decide whether the result is ready.
You might discover that you gave too little context. Or that the assignment was too broad. Maybe your assistant needs to ask you a question before it can move forward. Start small enough that you can notice those things and try again.
Once the result is what you want, you can make this an ongoing responsibility. Maybe you should make it a private draft each Friday. Or how about preparing your accomplishments before a review meeting, with an agreed trigger and enough lead time. The point is, you decide together which sources to use, when it should ask you, and what needs your review before anything is shared. Itβs your assistant, let it assist you!
You delegated, you didn't tell it how to do the work. You gave it enough context and permissions to get the results you wanted, and when you wanted them.
And again just a friendly reminder that you donβt need to be on a management track to delegate. Keep practicing. Pick a task to hand off, see what comes back, and improve it together. Don't define how it needs to do the work, let it be creative, and let it ask for help when it needs it.
Iβm working on a video about this, and Iβd love your help shaping it. Whatβs one thing youβd hand over to your assistant, and what would you want to check before trusting the result?
If you try the week in wins/brag journal exercise, what did you have to clarify before the result became useful? Did it feel good?
I hope this little experiment shows up up in your weekly wins, because in my book, learning is winning, and if you made it this far, you dedicated some real attention to this.
If you haven't yet found an AI assistant you love, I recommend taking dots for a spin. Dorothy and I worked through this post in my headphones on a lovely stroll through my neighborhood. Having an always on agent that I can call whenever an idea pops in my head has really brought my delegation skills to a new level. I hope it does for you too!
Keep me posted on your journey!
Originally published by Dev.to AI. Aggregated on AIWithGhost for educational purposes β full credit and traffic to the original publisher.

