The system that slowed down the "Code Girls" of WWII is the same one stalling your AI strategy today.
While reading Code Girls, I expected a story about unrecognized talent. What I also found was a cautionary tale about fragmentation. In the 1940s, the Army and Navy weren’t just fighting the Axis powers; they were fighting each other. Competing for talent and hoarding breakthroughs at a time when we needed speed and collaboration.
They were operating in silos. And eighty years later, we’re still building them.
It’s not just the lack of recognition for these women that bugs me. It’s how much progress was slowed not by a lack of talent, but by fragmentation, competition, and a system that wasn’t designed for collaboration.
When I say silos, I don’t just mean teams operating separately. I mean work happening in parallel without shared context.
Decisions being made without visibility across functions.
Progress that doesn’t build on what’s already been done.
The most obvious and pervasive example in the book is the divide between the Army’s Signal Intelligence Service and the Navy’s OP-20-G.
Both organizations were working on overlapping codebreaking challenges. Both had brilliant people. Both were making progress. But they weren’t fully working together. In fact, they were competing.
Information sharing was limited. Breakthroughs didn’t always move quickly between teams. Effort was duplicated.
And this wasn’t about bad intentions. The reward system drove this behavior. The groups battled over talent. They raced to be the first to share a discovery. They tackled the same problems instead of spreading the work across groups best suited for the challenge.
This wasn’t just competition. It was silos preventing progress at scale. In a historic time where speed mattered, fragmentation cost our nation greatly.
And we haven’t truly taken this lesson to heart. I’ve seen versions of this play out in my own work more times than I’d like to admit.
At one company, we had multiple teams creating content around the same core topic. Marketing was publishing thought leadership. Sales had their own decks. Product was building technical briefs.
All of it was good. None of it was aligned.
We weren’t sharing early. We weren’t building on each other’s work. We were solving the same problem three different ways.
The result wasn’t just inefficiency. It was diluted impact. We had more content… and less clarity. Not because of a lack of effort. Because we were operating in silos.
I’ve seen it with partners too. Two strong organizations, both saying they wanted to collaborate. But underneath that was a quiet competition over visibility, ownership, and credit.
Instead of combining strengths, we spent time negotiating positioning. What could have been a differentiated, unified message turned into a compromised one.
Not because we lacked expertise. Because we couldn’t fully let go of control.
Even in collaboration, we were still protecting our own silos.
I’m seeing this same pattern play out right now with AI. The difference is, AI doesn’t stay neatly inside one team.
When sales starts using AI, it has implications for compliance, legal, and customer experience. When marketing adopts it, it impacts brand, data usage, and how product is represented. When operations or product teams implement it, it affects technical architecture, security, and downstream workflows.
AI cuts across everything. But we’re still approaching it in silos. And with AI, silos don’t just slow things down. They create risk.
Sales is experimenting. Marketing is building workflows. Product is testing separately. Legal and compliance are often reacting after the fact. All of it happening at the same time. With very little coordination.
The result isn’t acceleration. It’s fragmentation.
We’re duplicating effort. We’re creating inconsistent experiences. And in some cases, we’re introducing risk because no one has a full view of what’s being implemented across the organization. AI has the potential to dramatically increase how fast we move. But without alignment, it does the opposite.
It amplifies the gaps. Different era. Same pattern.
We don’t have an AI adoption problem.
We have a coordination problem.
And if I’m being honest, I’ve been part of the problem too.
There were times I held onto an idea too long before sharing it. Times I wanted to fully form something before letting others in. I can look back today and see where this behavior has slowed me down.
It didn’t feel like competition in the moment. It felt like ownership.
But the effect was the same.
That’s why this particular theme in Code Girls is hitting me so hard.
The barriers weren’t just external. They were structural. And in some ways, behavioral.
And they’re not just historical. They’re still showing up today. The tools have changed. The silos haven’t.
That’s the opportunity. To reflect on the past and ask how we do better today.
I’ve been thinking a lot about the questions we ask. Maybe the better question isn’t whether we create space. It’s whether that space connects… or separates. Because even well-intentioned spaces can become silos if we’re not careful. We’ve built some really strong women-only spaces. And I understand why. They create access, confidence, and community.
But I think we need to ask a harder question:
If the goal is influence, impact, and progress… can we get there by operating separately?
I’ve experienced real value in these spaces. And I’ve also seen moments where frustration turns into distance. Where “we need our own space” quietly becomes disengagement from the broader system. That might feel better in the short term. But it doesn’t move the work forward.
The systems we’re trying to change aren’t made up of one group of people. They’re made up of all of us.
We don’t fix fragmentation by creating more of it. Even when it’s well-intentioned.
If I strip this down to what actually needs to change, it’s not abstract. Collaboration isn’t a mindset. It’s a set of behaviors:
None of this is complicated. But it does require a shift from “what’s mine” to “what moves this forward.” From protecting silos… to connecting them.
This is a design problem. If you’re in a leadership role, are you asking the right questions to prioritize collaborative, supportive culture?
And on an individual level, I’m asking myself this:
Where am I still working in a silo… even when I say I value collaboration?
This week, I’m choosing to act more collaboratively by looking for opportunities to:
If we actually want better outcomes, this is where it starts. Better outcomes don’t come from better silos. They come from better alignment.