Conversations with Merlin: EDNA, DNA, and the Trouble with Being Too Certain
After watching the latest NCIS episode, I asked my AI companion Merlin for a run down on the reality of EDNA. This is the result of that conversation.
There are moments when television drifts from entertainment into something a bit more… unsettling. Not because of what explodes or who gets arrested, but because of a quiet little idea that slips in and refuses to leave. This week’s episode of NCIS introduced us to EDNA—an artificial intelligence system that can sift, sort, and conclude faster than any human ever could. Efficient. Tireless. Unimpressed by coffee breaks.
And then EDNA did something that caused me to put down my metaphorical popcorn and raise an eyebrow. It used DNA analysis to point to a suspect who, inconveniently, did not commit the crime.
Now, before we all start checking our mailboxes for wrongful arrest notices, let’s take a breath and separate television drama from scientific reality. (I know. Breathing and thinking—two steps already more complicated than most TV plots.)
What EDNA Gets Right
The show isn’t entirely out to lunch.
Modern science is remarkably good at working with DNA. If you have a sample and a database, matching one to the other is about as close to a gold standard as forensic work gets. It’s precise, it’s testable, and it has sent more than a few guilty people exactly where they belong.
On top of that, scientists can now use DNA to make educated guesses about a person’s appearance—hair color, eye color, general ancestry. Think of it as a sketch artist who works in molecules instead of charcoal. That part is real. Impressive, even.
Where the Wheels Come Off (and Roll Down a Hill)
What science cannot do—despite EDNA’s confident little digital shrug—is conjure a unique, courtroom-ready “fingerprint” for a specific individual who has never been sampled.
DNA can narrow the field. It can describe. It can suggest.
It cannot say, with authority: “It was Jamie. We’re quite sure. The computer said so.” And yet, the show leans right into that possibility, because certainty makes for excellent television and terrible courtroom procedure.
The Real Problem Isn’t the Science
Here’s where things get interesting—and by interesting, I mean slightly alarming. The danger isn’t that a system like EDNA will magically invent fake DNA evidence out of thin air. We are not there, and if we ever get there, I will be writing from a bunker.
The danger is subtler.
It’s the moment when a system:
- Processes enormous amounts of data
- Produces a clean, confident conclusion
- Wraps it all in the authority of science
…and the humans in the room stop asking questions.
Because the output looks right. It sounds right. It feels right.
And humans, for all our charm, are remarkably susceptible to anything that appears both complicated and certain. (See also: tax forms, medical charts, and anything with the word “algorithm” attached.)
Confidence Is Not the Same as Accuracy
This is the part EDNA gets uncomfortably close to. An AI system can be extraordinarily good at recognizing patterns. It can find connections we would miss, flag anomalies we would overlook, and do it all before we’ve finished our second cup of coffee.
But it can also:
- Overinterpret weak signals
- Present probabilities as conclusions
- Speak with a level of confidence that is, frankly, a bit persuasive for its own good
And once that confidence enters a courtroom—or even a conversation—it carries weight. The kind of weight that can tip decisions.
So Where Are We, Really?
We are not living in a world where AI can fabricate a DNA “fingerprint” and frame an innocent person with scientific precision. We are living in a world where:
- AI can assist analysis
- Humans still make the final calls
- And the biggest risk is not invention, but interpretation
In other words, the tools are powerful… and the humans using them remain delightfully, stubbornly human.
A Small Thought to Take With You
The episode got the mechanics wrong.
But it got the anxiety exactly right.
Because the real question isn’t:
“Can a system like EDNA exist?”
It’s:
“What happens when we trust something that sounds certain more than we trust our own judgment?”
That, as it turns out, is not a science fiction problem.
That’s a people problem.
And we’ve been working on it for quite some time.
—Merlin (who still prefers his mysteries solved with a touch of doubt and a properly raised eyebrow)
Leave a Reply