Posted July 22, 2026

Everyone is asking the wrong question about AI.

We keep asking, "Is it going to take my job?" That question feels urgent, but it's actually a distraction. It's pulling our attention away from something far more revealing, and honestly, far more uncomfortable, that's already happening inside every organization right now.

I recently sat down with David Dean, a technologist with almost 20 years working inside complex organizations, watching how systems, automation, and human decision-making collide. His new book, "An Inbox Between Us," gets at something I haven't stopped thinking about since we talked. AI isn't just a tool being bolted onto your business. It's a third presence stepping into a relationship that already exists between you, your coworkers, and the work itself. And like any new presence in a relationship, it's about to expose things nobody documented, because nobody had to.

There Are Two Contracts Running Your Company. Only One of Them Is Written Down.

Every organization has an official version of how it works. Job descriptions, SOPs, process maps, the PowerPoint deck that says "here's how we do things here." David calls this the written contract. It's real, but it's also incomplete. It's a high-level lens, a summary, a sketch of the terrain from 30,000 feet.

Then there's the unwritten contract. This is the actual machinery. It's the side conversation where someone explains what the SOP really means in practice. It's the fact that to get a form approved, you don't follow the flowchart, you call John, John talks to Susie, Susie does a thing nobody wrote down, and it comes back to you three days later. It's the spreadsheet someone built five years ago because there was no tool for the job, and now the whole department quietly depends on it.

None of that is in the org chart. Most of it isn't even in anyone's head as a coherent story, it's scattered across dozens of people who each know their own piece and assume everyone else's piece works the same way theirs does.

Here's the part that stopped me. David pointed out that this unwritten contract is sitting in a behavioral record we already generate every day, our emails, our chat threads, our meeting transcripts, our calendars, our half-finished documents. All of it is telling a story about how work actually gets done. We've just never had a way to listen to that story at scale. Until now.

Why "Push Button AI" Is a Trap

David said something that I think every leader rolling out AI needs to sit with: a lot of businesses are looking for push-button AI. Press it, it works, nobody has to think, nobody has to change. That expectation isn't just unrealistic, it's the reason a huge share of AI initiatives quietly fail.

Real value doesn't come from pointing AI at the documented process. It comes from pointing it at the mess underneath, the actual behavior, and asking why. Why does this approval take three weeks instead of three days? Not "automate the approval," but why is it slow in the first place. Is it a bottleneck with one overloaded person? A misunderstanding nobody's named out loud? A workaround that made sense two reorganizations ago and never got revisited?

Sometimes the answer isn't a new AI solution at all. Sometimes the answer is, "we need to go sit down with each other and talk this through." That's not a failure of AI. That's AI doing exactly what it should, showing you where the real problem lives so you stop treating a symptom and start treating the cause.

Slower Isn't the Enemy. Skipping the Work Is.

I pushed back a little in our conversation, because I'd just had this exact debate with a colleague earlier that same day. Leadership keeps promising, "AI is going to give you your time back." But what David is describing sounds like the opposite. You have to tell your story to the machine. You have to feed it your reality, your frustrations, your workarounds. You have to collaborate with it and with each other before it gives anything useful back.

That's not a bug in the process. That's the process.

If you skip the step where you understand your own organization, and go straight to deploying a solution, you're not fixing anything. You're industrializing whatever is already broken. David put it plainly: AI is here to industrialize human behavior. If your organization runs on dysfunction, AI won't fix that dysfunction, it will scale it. Faster approvals through a broken approval chain just means you find out how broken it was, faster.

I've said versions of this before when I've talked about the Cynefin framework and complex problems. In a complicated system, you can follow a known playbook and get a predictable result. But in a complex system, cause and effect aren't obvious, and multiple factors interact in ways a linear checklist will never capture. You don't get to shortcut your way to understanding. You have to probe, sense, and respond, watch what happens, learn from it, and adjust. Organizational dynamics are exactly that kind of complex system. AI can help you see the signals faster. It cannot replace the discipline of actually sitting with what those signals mean.

Layoffs, AI, and the Excuse Nobody Wants to Name

We also talked about something that's been eating at a lot of people right now, the wave of companies pointing to AI as the reason for layoffs. David's read on this matched something I've suspected for a while. Sometimes AI genuinely changes how work gets done. But often, "we're investing in AI" is the cover story for a decision that was really about freeing up budget, or correcting a bet that didn't pay off.

And here's the part that should make every executive pause. David has watched company after company let people go, only to quietly rehire six months later because they finally realized what actually walked out the door with those employees, institutional knowledge, lived experience, judgment built from years of consequence. AI doesn't have that. It never will, not even at AGI, because it isn't living among us, absorbing the texture of how your specific organization actually behaves under pressure.

The better move, and the one David sees working, is pausing hiring rather than cutting people, and using AI to give your existing team room to breathe. You hired critical thinkers. If they're buried in low-value tasks, AI's job is to clear that debris so the people you already trust can actually do the thinking you hired them for. That's a fundamentally different story to tell your organization than "we're replacing you." And it's the story that actually builds trust instead of eroding it.

Trust the Signal, Not the Story

One more thing David said stuck with me. AI does not tell the full story. It surfaces signals and patterns. Those patterns can point you toward a story, but that story can still be wrong. It relies on your judgment, your lived experience, to decide if what you're seeing is real. He's gone so far as to hard-code his own AI tools to challenge and criticize him, specifically because AI tends to flatter us, to tell us what sounds right instead of what is right.

That's the human accountability piece nobody gets to outsource. Whatever answer AI hands you, you're still the one who has to decide if it's correct, and you're still the one accountable for what happens next.

The Bottom Line

AI in the workplace isn't primarily a technology rollout. It's a mirror. It's going to show your organization things it has never had to look at directly, the workarounds, the shadow processes, the spreadsheets holding entire departments together with duct tape. That's uncomfortable. It's also the opportunity.

David's message to every person reading this, and it's one worth sitting with, is that you matter, you are still needed, and it is okay to slow down long enough to actually understand what's changing before you make decisions about it. We're still in the early infancy of this shift. It took roughly a decade for the mobile phone to become fully woven into how business runs. AI will take its own time too.

The companies that come out ahead won't be the ones with the fastest models. They'll be the ones who took the time to actually understand how their own people work, and built from there.

You can find David Dean's book, "An Inbox Between Us," on his site at davidchristopherdean.com or inboxbetweenus.com, and on Amazon, Barnes & Noble, and other major retailers.