Why you should keep your AI tab open all day
The most common AI mistake I see is closing the tab.
Most nonprofit staff use AI like a reference book. They open it, look something up and shut it again. Every task starts from zero, and the AI only sees the small piece you choose to show it. That keeps you in a linear way of working, where nothing happens until you stop what you’re doing and ask.
The fix is a habit. Keep your AI assistant open all day, on a second screen or in its own window, and talk to it as you work.
Why an open tab changes how you use AI
When AI is something you open for a big task, you only think of it for big tasks. A lot of the time savings are hiding in the small ones, like the list you tidy up before a meeting or the two spreadsheets you compare by eye.
So start your day by telling it what’s ahead. Something like this:
Here’s my to-do list for today. I’m working on a donor appeal, a board report and cleaning up a volunteer list. As I go, I’m going to tell you what I’m doing. Point out where you can save me time.
With the conversation open all day, AI starts to feel more like a teammate who knows what you’re working on.
Narrate your day
The next habit is narrating your work. We often don’t notice which tasks are shovelling until someone points it out. Most of us don’t have a full-time assistant sitting next to us, so let the AI play that role.
Say what you’re about to do, out loud or in the chat. For example: “I’m about to spend 20 minutes cross-referencing these two spreadsheets to find duplicate donors.” A good assistant will come back with something like: “I can do that for you in about ten seconds if you upload both files.”
Before you upload anything with donor details, make sure the tool is one your organization has approved for that kind of data.
Narrating is how you move from guessing what AI can do to knowing what it can do. It’s the bridge between using a chatbot now and then and having AI built into your daily work.
From habits to agents
Narrate for a few weeks and you’ll start to see patterns. You’ll catch yourself thinking, “Wait, I ask it to clean up this exact report every Wednesday.”
That’s when it’s time to build an agent. An agent is a habit you’ve turned into a process that runs by itself. Instead of narrating the task each week, you set things up so the AI watches for the data and does the work without being asked.
- Linear. You find the data, hand it to the AI and copy the result back.
- Parallel. The agent sees the new data, cleans it and lets you know when it’s done.
At that point you’re managing a small digital workforce.
The habit is about to get a big upgrade, too. Newer desktop tools like Claude Cowork are moving AI out of the browser and onto your computer. The old friction was uploading files and pasting text back and forth. Cowork can see your screen and work in it. Say you’re in your CRM, updating donor fields by hand from an email. A tool like that can take over the updates while you jump on a call.
How to build the habit, one rung at a time
I think of AI productivity as a ladder. Here’s how I’d climb it:
- Keep it open. Put your AI assistant in its own window and leave it there all day. Give it your to-do list first thing, so it has context for everything that follows.
- Narrate as you go. Before you start a task, tell the AI what you’re about to do and how long you expect it to take. Ask it to speak up if it can handle part of it faster.
- Keep a repeat list. Any time you ask for the same thing twice in a week, write it down. After a month, the list shows you which tasks are worth automating.
- Turn the top repeat into an agent. Pick the task on your list that’s most predictable and lowest risk. Set it up to run on its own and report back, and check its work closely for the first few weeks.
- Bring in a desktop tool for the hand-offs. Once you trust an agent with one task, look for the spots where you’re still moving data between screens by hand. That’s where desktop tools help most.
For leaders: help your team find the first rung
You can’t jump to the frontier if your team hasn’t built the basic habit yet. Agents and desktop tools build on the habit of keeping AI open and talking to it, so that’s where to start.
Most nonprofits I meet are still looking for the ladder. Helping teams climb it is a big part of my strategy and training work. A few questions to bring to your next team meeting:
- Who keeps AI open all day, and who only opens it for big tasks?
- What did people narrate this week that AI could have handled?
- Which tasks do we repeat every week that could become an agent?
- Do staff know which AI tools are approved for donor and client data?
Then pick one person to try the open-tab habit for a week and report back on what they found.