Interview with the Vampire is really about an unreliable knowledge base

I started watching because I remembered Tom Cruise. I ended up thinking about organisational memory.

I watched Interview with the Vampire because I remembered the Tom Cruise film. Quite vividly, actually.

Tom Cruise as Lestat. Brad Pitt as Louis. Kirsten Dunst as Claudia. That gothic, slightly ridiculous, completely compelling version of New Orleans and Paris. I hadn’t watched it for years, but it had stuck.

So when I started the newer television adaptation, part of the attraction was simple curiosity. What would they do with something I already had such a strong memory of?

Quite a lot, as it turns out.

And somewhere along the way I became less interested in the vampires than in the interview.

Because Interview with the Vampire is a surprisingly good story about what happens when somebody tells you a story — and you don’t entirely believe them.

The interviewer matters

Daniel Molloy could easily have been a framing device. Louis talks. Daniel records it. We get the vampire story.

Except Daniel doesn’t behave like that.

He’s older than when he first encountered Louis. More sceptical. Less impressed. And crucially, this isn’t the first time he’s heard the story. He has material from before.

So he does what a good interviewer should do. He listens. He compares. He notices when things don’t fit.

Why are you telling it differently now?

What happened here?

Why did you leave that out?

Are you remembering this incorrectly — or choosing to remember it this way?

Suddenly the interview isn’t the frame around the story.

It is the story.

Memory isn’t evidence

That’s probably the part that hooked me.

I’ve spent a large part of my working life interviewing people, and one thing you learn fairly quickly is that an interview is not an information retrieval exercise.

People aren’t databases.

Ask somebody what happened and you don’t receive an objective record of events. You receive their interpretation of it. Memory gets mixed up with emotion, hindsight, loyalty, embarrassment, self-preservation and everything that has happened since.

That doesn’t mean anybody is lying. It means a story told by a human being is inevitably a constructed thing.

Interview with the Vampire makes that construction visible. Louis is simultaneously protagonist, witness and narrator. Daniel has to work out how much confidence to place in the narrator while using the available evidence to reconstruct what might actually have happened.

And then Armand complicates everything further.

Having the material isn’t enough

This is where my work brain switched on.

Daniel has recordings. Testimony. Notes. Memories. Access to the people involved. In modern terms you might say he has an extremely rich knowledge base.

None of it automatically produces understanding.

The important work happens when he starts connecting things. A contradiction here. An omission there. Something Louis said decades ago that doesn’t quite match what he’s saying now. A detail that seemed insignificant until another piece of information changed its meaning.

The material existed. What was missing was the interpretation.

That gap is most of what I find interesting about Editorial Intelligence. We’re entering a period where retrieving information is becoming extraordinarily easy — AI can search enormous bodies of material, summarise interviews, transcribe meetings and surface things we had forgotten. That is genuinely useful.

But retrieval isn’t understanding. Somebody — human, machine, or some combination — still has to ask what it actually means.

The knowledge base is also a narrator

Here is the part I keep returning to, and it’s the reason this stopped feeling like an analogy.

It’s tempting to say Daniel’s problem doesn’t apply to organisations. Louis is unreliable in a way a database isn’t. A repository doesn’t protect its reputation, or quietly reshape a memory it finds painful.

Except most of what an organisation knows didn’t arrive as data. It arrived as accounts.

Customer stories are written to be flattering. Win and loss notes are recorded by the person who won or lost. Launch retrospectives are shaped by who was in the room and what they needed the conclusion to be. Research is commissioned with a hoped-for answer rather more often than anybody admits.

None of that is dishonesty. It’s what happens when people record events they were part of.

Which makes Daniel’s job a fairly literal description of the work rather than a metaphor for it. Anybody trying to establish what an organisation actually knows is not reading a record. They are interviewing a narrator with a stake in the story.

Where the comparison breaks

Daniel has advantages nobody working with organisational memory gets.

His narrator is alive and in front of him. He can ask again. He can put a contradiction to Louis’s face and watch what happens next, which is the single most useful thing an interviewer can do.

Organisational memory rarely offers that. The person who wrote the note has left. The context they assumed everybody shared is gone. The contradiction is visible and unresolvable, because there is nobody left to ask.

That makes the interpretive work harder, not easier — and it’s the strongest argument I know for capturing the reasoning at the time rather than only the conclusion. The record of what was decided survives. The record of why it was decided is what you’ll actually want in two years, and it’s the part nobody writes down.

Perhaps that’s why Daniel interests me

He isn’t the most spectacular character in a programme full of spectacular characters.

He can’t fly. He doesn’t live forever. He isn’t especially interested in being seduced by any of it.

He just keeps asking awkward questions.

And I increasingly think that’s an underrated skill. Not gathering more information. Not accepting the first coherent narrative you’re given. But going back through the material and asking whether the story you’ve constructed actually survives contact with the evidence.

I started watching because I remembered Tom Cruise.

I ended up watching a journalist interrogate a knowledge base.

Which probably says something about me.