The case for corporate amnesia

Most organizations worship institutional memory. They build knowledge bases, write handoff documents, run onboarding programs that take months — all in service of the idea that accumulated context is an asset. The longer someone has been around, the more they know. The more they know, the more valuable they are.

What if that’s backwards? What if what they’re actually accumulating isn’t knowledge but assumptions? Stale models of how things work that haven’t been tested against current reality in months, maybe years. The senior person who “knows where everything is” often knows where everything was. And the organization rewards them for it because the alternative — admitting that their mental model might be wrong — feels like starting over.

What I think is actually happening in most organizations: institutional memory is a mix of about 20% genuine expertise and 80% accumulated habit. And nobody ever audits which is which.

Think about a hospital. A veteran nurse on a ward has been there for twelve years. She knows which doctor prefers which suture kit, where the extra blankets are stored since the renovation, how to calm a specific type of anxious patient. That’s real knowledge. But she also “knows” that the lab always takes 45 minutes on a Thursday, that the new resident can’t be trusted with intubation, that the pharmacy messes up orders from the night shift. Some of that was true once. Some of it was never true. All of it shapes every decision she makes, every day, without being examined.

Now imagine you could give that nurse perfect amnesia about everything except the building layout and the medical protocols. She’d walk onto the ward tomorrow and discover the current state of things. The lab’s turnaround has improved. The resident is actually competent. The pharmacy has a new system. She’d be working with reality instead of working with a twelve-year palimpsest of reality.

You’d never do that, obviously. The cost of her genuine expertise is too high. But the thought experiment reveals something: what we call “experience” is a package deal. You get the wisdom and the baggage in one bundle, and you can’t easily separate them.

Organizations have been built around the assumption that this bundle is mostly wisdom. The entire concept of seniority rests on it. Promotion structures, mentorship programs, the way we weight opinions in meetings — all of it assumes that accumulated context trends toward truth over time.

The opposite might be closer to accurate. Accumulated context trends toward comfortable fiction. You stop checking your assumptions because they’ve been working — or at least, you think they’ve been working, because nobody’s measured.

This is why new hires see things veterans don’t. Not because new hires are smarter. Because they haven’t accumulated the comfortable fictions yet. Every veteran knows this. They just call it “they don’t understand the culture yet” instead of “they haven’t learned to ignore the problems yet.”

There’s a version of this in how teams communicate. A team that’s worked together for three years develops shorthand. They stop explaining context because everyone “already knows.” The problem is that what everyone “already knows” slowly drifts apart. Each person’s mental model evolves independently through private experiences — conversations they had alone, problems they solved without telling anyone, assumptions they updated quietly. After three years, the team thinks it has shared understanding. What it actually has is six different mental models that were synchronized once and have been diverging ever since.

The fix for this in high-stakes environments is the briefing. Military units don’t assume shared context. They rebuild it before every operation. A surgeon doesn’t assume the team remembers the plan from the morning — they run the checklist again. Airlines don’t let pilots rely on “I’ve done this route a hundred times.” They read the approach procedure every single time.

These fields figured out something most knowledge work hasn’t: fresh context, loaded on demand, beats accumulated memory. Not because memory is worthless, but because memory is unreliable, and unreliable memory that everyone trusts is more dangerous than no memory at all.

We’re entering a period where organizations can actually choose between these two models. You can have workers — human or otherwise — who accumulate context over time, building up history and relationships and judgment. Or you can have workers who start fresh every time, loading only what’s relevant to the task at hand from authoritative sources.

The instinct is to say the first model is obviously better. Of course you want experience, context, relationships. But ask yourself: how much of what your most experienced people “know” have they verified in the last six months? How much of your organization’s institutional knowledge is actually institutional habit?

The second model — fresh eyes, loaded context — has a property that feels counterintuitive but turns out to be powerful: it can’t lie to itself. It has no accumulated assumptions to protect. It reads the current state of things because it has nothing else to read. When the procedure changes, it sees the new procedure. When the situation shifts, it sees the new situation. It doesn’t have to unlearn anything because it never learned anything in the first place.

This isn’t theoretical. Anyone managing a team knows that the hardest conversations aren’t about competence. They’re about updating mental models. “I know you’ve been doing it this way for years, but the situation has changed.” That sentence accounts for more organizational friction than any lack of skill. People don’t resist change because they’re lazy. They resist it because their identity is wrapped up in what they know, and updating what they know feels like losing part of themselves.

What if you could build an organization where that sentence never needs to be said? Not because nothing changes, but because no one’s identity depends on any particular way of doing things?

The obvious objection is that you lose relationships. You lose the judgment that comes from watching the same customer, the same market, the same team dynamics play out over years. Fair enough. But I’d push back: how much of that judgment is actually good? Have you tested it? Or have you just assumed it’s good because the person who has it has been around a long time?

The really provocative implication here is about what we should value in people. If accumulated context can be loaded on demand — and increasingly it can — then the premium shifts. The valuable person isn’t the one who remembers the most. It’s the one who makes the best decisions with whatever context they’re given. Decision quality, not context volume. And those are very different skills.

Some people are extraordinary at accumulating and retaining context but mediocre at deciding. We’ve been promoting them for decades because in a world where context was scarce, having a lot of it was valuable. That world is ending. Context is about to be cheap. Judgment is about to be expensive.

And the organization that figures this out first — that stops rewarding memory and starts rewarding the ability to decide well under fresh conditions — has an advantage that compounds. Because everyone else is still paying the tax on their institutional memory: the comfortable fictions, the divergent mental models, the sacred cows that haven’t been examined since 2019.

There’s one more piece of this that matters. The person who coordinates — who sits in the middle and sees across all the moving parts — that person does need memory. Not because memory is inherently good, but because coordination requires a persistent model of who’s doing what and why. You can’t coordinate fresh every time. The coordinator is the exception.

But notice what that means: out of an entire organization, exactly one function needs persistence. Everything else is better off forgetting and reloading. The ratio isn’t fifty-fifty. It’s not even close.

So the wrong question is “should organizations value institutional memory?” The right one is: what is the smallest amount of persistence you can get away with, and are you sure you’re not carrying more than that?

Because most organizations carry far more than that. And they call it culture.