The Fire Service Needs a Second Brain
Robert Grand · Battalion Chief who still runs calls
My city computer is locked down tight. For a long stretch, Chrome wasn’t even an allowed browser on it. That came back recently. The one AI tool I’m sanctioned to use on that machine is a restricted version of Copilot, scoped and fenced about the way you’d expect a municipality to scope it. None of that is a complaint. Every one of those restrictions exists because somebody’s actual job is to keep a city from getting taken down, and that threat isn’t theoretical.
So I do my best thinking somewhere else.
Nobody Locked That Machine Down for Fun
It would be easy to write this as a story about IT getting in the way, and that’s not the story. Cities are targets. Municipal networks get hit, and when they do it isn’t an inconvenience, it’s a public safety event. The fire department is one department on a network carrying payroll, utilities, courts, and everything else. The person who wrote those restrictions was doing their job correctly. If I were them, with their risk on my desk, I’d have written most of the same rules.
The restrictions aren’t aimed at me. They still landed on me, and I responded the way people respond, which is that I found another way to do the work.
I use my own machine. Not to get around anybody, and not for anything touching protected information. For the thinking. The tools I want to think with aren’t available inside the wall, so the thinking migrated to where the tools are. There’s no rebellion in it. It’s water finding the low spot.
It’s a Thought Partner, Not a Ghostwriter
Here’s where I lose some people in this service, so let me be specific about what I actually do with it.
I’m not generating documents. The fear in Fire & EMS, and it’s a reasonable one, is that this technology means somebody hands up a poorly formatted memo full of confident language that nobody wrote and nobody read. I’ve seen that output. It’s obvious, it’s bad, and it earns the reaction it gets.
What I’m doing is different. I open it and I argue with it. I put a half-formed idea in front of it and ask what I’m missing. I take something that’s about to go up the chain and work it over until it’s tighter and the weak part is exposed. It doesn’t do my job. It makes me faster at the part of my job that happens between my ears, which as a Battalion Chief is most of it.
That’s the distinction the whole conversation keeps missing. The question isn’t whether a machine can write your report. It’s whether you have anything to think with when you’re building an argument you’ll have to defend in a room full of people with more rank than you.
It’s open every day and I go there almost before anywhere else. I’ve turned that over a few times, wondering whether leaning on it is making me sharper or just faster. My read, so far, is sharper. I keep checking, because that’s the kind of thing a person should keep checking.
It Passed Security and It Failed Me
The restricted tool on my city machine isn’t there because somebody made a bad decision. It’s there because it cleared the bar it was asked to clear. It met the security parameters. That was the requirement, it was a legitimate one, and it got satisfied.
Nobody was ever asked whether it worked for the person doing the job.
That’s the gap, and it isn’t unique to one tool or one city. When a department picks technology, the questions that get asked are the ones somebody is accountable for answering. Does it meet security. Does it fit the budget. Will the vendor support it. Those all matter and they all have an owner in the room. The second question usually doesn’t have an owner: does this make a battalion chief better at their job at three in the afternoon. Score only the first set and you get a tool that’s compliant and unused, and people working around it on their own equipment.
That’s not a hypothetical. That’s my Tuesday.
The Kitchen Table Is Already Doing This Research
I’m not unusual, and that’s the part worth paying attention to.
Fire Engineering ran an honest piece on how this technology is actually entering our service, and named something most of us have watched happen without saying out loud: there are two speeds running at once (Fire Engineering, https://www.fireengineering.com/firefighting/fire-leadership/from-the-firehouse-to-fireground-how-ai-is-reshaping-the-fire-service/). The official speed is cautious and organizational and moves at the pace of formal buying, which is slow and narrow. The unofficial speed is individual, decentralized, and already running, at kitchen tables, on personal phones, with tools anybody can open in ten seconds. FireRescue1 made the companion argument, that this supports the fire service only if safety leads the conversation instead of trailing it (FireRescue1, https://www.firerescue1.com/artificial-intelligence/ai-can-support-the-fire-service-but-only-if-safety-leads).
Both are right. Both describe a service where adoption stopped being a decision the organization makes and became something it finds out about later.
Somebody on your department has already used one of these to rough out a lesson plan, make sense of a code section, or tighten a draft they were stuck on. Probably several somebodies. Probably good ones. Probably without mentioning it, because nobody ever said whether it was allowed, and in this service the safe play on a question nobody has answered is to not raise your hand.
The Workaround Is the Map
This is where I’ve landed, and it took me a while to get here.
When a department’s people start routing around the sanctioned setup to get work done, the instinct is to treat it as a control problem. Write the policy. Close the gap. Get everybody back inside the wall. I understand that instinct and I don’t think it’s crazy.
I think it reads the signal backwards.
Every one of those workarounds is a person telling you, on their own time and with their own equipment, exactly where the job has been grinding them down. Nobody goes to that trouble for a task that was already easy. The tasks people quietly reach for help with are a map of the administrative load this service has been stacking on company officers and battalion chiefs for twenty years, drawn by the only people qualified to draw it. That’s not a violation to catch. That’s free research the department didn’t pay for and doesn’t collect.
So the first move isn’t a policy. It’s a question, asked so people believe it’s safe to answer honestly: what are you already doing, and what were you trying to make easier? If the honest answer gets somebody counseled, you’ll get the other answer, and you’ll have burned your one shot at a real inventory.
What a Department’s Second Brain Would Actually Know
Here’s the part that gets me, and it’s the reason I keep coming back to this.
What I have is a second brain for me. It knows my context because I feed it my context. It knows how I think about a staffing problem, what I’ve been reading, what I’m working on this month. That’s useful and it’s also completely bounded, because when I retire, it retires. Everything it learned about how I do this job walks out with me, the way it always has.
The fire service doesn’t have the other thing. Nobody has a second brain that belongs to the department. Nothing in our world learns. We run CAD, records systems, ePCR platforms, preplan software, and every one of those stores. Storage is not memory. Not one gets smarter about our district because we ran a thousand calls in it, and nobody is training anything on their own after action reviews.
That isn’t a failure, it’s where the field honestly is. It does mean something worth sitting with. The most valuable thing a fire department owns is what its people know about operating in their own city, and right now that knowledge lives in individual heads and on personal laptops. Mine included.
And there’s a question sitting underneath all of it that I don’t think we’ve asked out loud yet. If a department does build this, who keeps it? If I leave, does what I taught it walk out with me, or stay with the department that paid for it? If we move platforms in six years, does any of it come along? Those aren’t technical questions. Those are doctrine questions, and doctrine is something a department writes for itself.
Which is what governance actually is, and I think we’ve let that word get away from us. It isn’t a purchasing rule and it isn’t a policy telling people what they’re not allowed to open. It’s a department deciding in advance, in its own words, what its tools have to do, what happens to what it knows, and who answers for it. Write that down and you measure what a vendor brings you against your own standard. Skip it and you measure it against theirs, which is the only standard in the room by default.
The technology to think with is here, and a real chunk of this service already figured that out on their own time, outside the wall, without asking anybody. What we haven’t built is the version that belongs to all of us. I’d rather we author that on purpose than wait and let it arrive as a checkbox on somebody’s renewal.
Robert Grand is a Battalion Chief at Eugene Springfield Fire with 24 years of service. He writes Frontline Intelligence, a newsletter on operational doctrine, technology, and leadership in Fire & EMS.
If this landed, share it with someone in Fire & EMS who needs to hear it.
Subscribe
From the floor, not the vendor booth. Two times a week.