markdown formatted blog content
The Cipher That Haunted America for 67 Years
Okay, folks, I need to tell you about something that's got the true crime world absolutely buzzing, and honestly? It kind of blew my mind when I first read about it.
The Zodiac Killer. You know the one—he terrorized Northern California in the late 1960s, sent cryptic messages to newspapers, and never got caught. The guy became so legendary that David Fincher made an entire movie about him. And here's the thing: he's one of the most famous serial killers in American history, and we still don't definitively know who he was.
Until now? Maybe?
Meet Alex Baber: The Guy Who Got Obsessed
So there's this 50-year-old guy from West Virginia named Alex Baber. He used to work as a bodyguard outside Charleston. Not exactly the background you'd expect for someone who might solve one of America's greatest criminal mysteries, right?
But Baber watched the movie Zodiac back in 2007 and became completely fixated on cracking the code. Specifically, he wanted to solve Z13—the shortest of the four ciphers the killer left behind. We're talking just 13 characters here, prefaced by "My name is..."
Thirteen characters. That's it. The whole thing.
So what did Baber do? He taught himself forensic science and cryptography using college textbooks. He learned to code. And then, a couple years ago, he started experimenting with AI.
Here's Where It Gets Wild
Baber prompted his AI program to generate potential names that would fit the cipher structure. You know what it spit out?
Seventy. Million. Possibilities.
I don't know about you, but my brain can't even comprehend that number. But here's the clever part—Baber didn't just sit there and read through 70 million names like some kind of masochist. He cross-referenced them against historical records like social security numbers and birth certificates.
And here's what they found: about 93% of those names belonged to people who never actually existed. They were fabricated, made up, or just statistical ghosts.
Once you eliminate those, and then filter out anyone who didn't match the Zodiac's physical description (white, middle-aged, around six feet tall), you're left with something manageable.
The answer? Marvin Merrill.
The Black Dahlia Connection Will Give You Chills
Now here's where this story takes a turn that honestly sounds like something out of a movie.
That name—Marvin Merrill? It turned out to be an alias used by Marvin Skipton Margolis.
Marvin Margolis was a suspect in the murder of Elizabeth Short.
You might know her by a different name: the Black Dahlia.
If you don't know this case, let me give you the quick version. Elizabeth Short was a 22-year-old aspiring actress found dead in a Los Angeles park in 1947. She'd been brutally mutilated, her body cut in half. Her dark hair and striking appearance earned her that haunting nickname.
It was one of the most infamous murder cases of the 20th century, and it was never solved.
Now, I don't know about you, but the idea that the Zodiac Killer and the Black Dahlia murderer might be the same person—or at least connected—sends actual shivers down my spine. These are two of the most legendary unsolved cases in American true crime history. The possibility that they might be linked?
Crime reporter Michael Connelly, who hosts a podcast about Baber's theory, told the New York Times: "I have no doubts."
That's a pretty bold statement.
But Here's What the Experts Say
Okay, now I need to pump the brakes a little and bring in some balance—because not everyone is convinced.
Klaus Schmeh is a German computer scientist who's written more than a dozen books about cryptology. His take? Z13 is probably unsolvable.
"The message is just too short," Schmeh told Popular Mechanics. "There are just too many possibilities."
That's a fair point, honestly. With only 13 characters, there are an almost infinite number of possible solutions. How do you really know you have the right one?
Schmeh isn't dismissing AI entirely, though. In fact, he says something that really stuck with me:
"At the moment, cipher solutions are falling like ripe fruit from the trees. The autumn of 2026 will undoubtedly go down in the history of cryptology. Thanks to AI-assisted methods, cryptographic cold cases that have resisted solution for decades are suddenly being cracked one after another."
Other Cases Are Getting Cracked Too
It turns out Baber isn't alone in this. AI is proving useful in all kinds of historical code-breaking situations.
Take Eren Bozbulut, a 22-year-old cybersecurity researcher from Turkey. He used AI to crack encrypted messages that a German spy named Walter Koehler left behind during World War II. The AI found the key: a single page from a Dutch prayer book published in 1893.
Or consider Anton Zaytsev, a Russian engineer who cracked a Cold War-era message found inside a glass bottle in Kaliningrad. The AI identified the key as a popular song from East Germany—specifically one beloved by a socialist youth organization.
And it's not just about old ciphers. AI is showing promise with ancient writing systems too. Google's DeepMind has a program called Ithaca that's helping scholars interpret damaged Greek inscriptions by predicting missing text. There's also work being done on Linear A—that mysterious writing system the Minoans used about 3,000 years ago that we've never been able to fully decipher.
The potential here is honestly kind of staggering. Imagine what we might learn about ancient civilizations if we can finally read what they wrote!
The Catch (Because There's Always One)
Elonka Dunin is a cryptology expert and author, and she's offering some crucial perspective:
"AI is remarkably good at making things up. It hallucinates and fills in blanks."
Ouch. But she's right. We've all heard about AI chatbots confidently stating incorrect information as if it were gospel truth. The same problem exists in code-breaking.
Dunin puts it simply: "A solution that looks right isn't necessarily a solution. You have to be able to demonstrate how you got there."
So while the results might be exciting—and they are exciting—there's still the crucial step of verification. You can't just trust the output. You need to understand the process and prove it holds up to scrutiny.
What I Think About All This
Here's my two cents, for what it's worth.
I think what Alex Baber did is genuinely impressive. This is a guy with no formal training who got obsessed with a cold case and taught himself multiple disciplines to chase down a theory. That kind of determination and curiosity is exactly the kind of thing that moves the world forward.
Whether he actually solved it? That's a different question. The brevity of Z13 really does make it incredibly difficult to verify. There might never be enough evidence to confirm or deny his solution with certainty.
But here's what I find most hopeful about this story: AI isn't going to replace human investigators. It's going to be a tool that enables them to work faster and explore more possibilities than ever before. Baber didn't just sit back and let the AI do everything—he guided it, filtered its results, and applied his own critical thinking.
And the broader applications are incredible. Ciphers from World War II getting cracked. Ancient languages potentially coming within reach. Historical mysteries that have puzzled scholars for generations finally yielding their secrets.
That's pretty amazing when you think about it.
Whether Marvin Merrill is the Zodiac Killer? We might never know for certain. But I think we're entering a new era of investigation—and personally, I can't wait to see what gets cracked next.