AI Didn't Give Me a Strategy. It Helped Me Find One I Already Had.
I'd known about Community Reinvestment Acts for years.
Not vaguely — specifically. I understood what they were, why they existed, which institutions had obligations under them, and why those obligations made them a natural fit for organizations doing reentry work. I had relationships in the banking sector. I had a clear picture of what my organization needed.
What I didn't have was the bandwidth to think it through.
That's not an excuse. It's just what the grind does. You carry information for months, sometimes years, without ever having the room to sit down with it and build something. The knowledge is there. The connections are there. The strategy is half-formed somewhere in the back of your mind. But the space to actually work it out — that's what's missing.
I stopped treating AI like a search engine around that time.
I'd been using it to generate outputs — drafts, summaries, templates. Useful, but limited. What shifted was bringing it a real problem instead. Not "tell me about CRA-aligned funding" but something closer to: here's what I'm trying to build, here are the constraints I'm working inside, here are the relationships I have and the ones I don't, here's what I'm afraid of getting wrong. Help me think.
That conversation is where the coalition strategy came from.
Not because AI knew something I didn't. It didn't have my relationships. It didn't know the specific people in the banking sector I could call, or the history I had with them, or the credibility I'd built over years of doing this work. What it did was hold the complexity long enough for me to see it clearly. It asked questions that surfaced what I already knew. It reflected the pieces back in an order that made the path forward visible.
The relationships still had to come from me. The vulnerability about my pain points — what I was really worried about, what I wasn't sure I could pull off — that had to come from me too. AI can't build a banking coalition. But it helped me think clearly enough to build one myself.
That distinction matters more than people realize.
Most of the conversation about AI in nonprofits focuses on what AI can produce. The grant draft. The donor email. The board summary. Those things are real and they save real time. But the more significant shift — the one that changed how I lead — was learning to use AI for thinking, not just producing. Bringing it the problems I hadn't had room to solve. Using it the way you'd use a very intelligent thought partner who's available at eleven on a Tuesday night when the rest of your team is asleep. The best thinking I've done in the last two years hasn't been solo.
It never really was — I've always thought better in conversation than in isolation. But before AI, that kind of thinking required another person in the room. Someone with enough context to follow the thread, enough patience to let the idea develop, enough honesty to push back when something didn't hold. That combination is rare, and it isn't always available when you need it.
AI made it available.
Not as a replacement for the people I think alongside. But as a way to show up to those conversations already clearer. Already having worked through the first layer of confusion on my own. Already knowing what I was actually trying to say.
The banking coalition came out of that. So did a clearer funding strategy, a sharper sense of which relationships to prioritize, and a confidence going into the room that I hadn't had before. Not because AI gave me something I didn't have. Because it helped me find what I'd been carrying without knowing how to use it.
That's what the Genies inside HeadspaceGenie were built to do — not just produce output, but help leaders think. Because the strategy is usually already there. Sometimes you just need the right conversation to find it.
