AI Thinks It Cracked Kryptos. The Artist Behind It Says No Chance

Over 35 years, amateur and professional cryptographers have tried to crack the code CryptosA glorious sculpture sitting behind the CIA headquarters in Langle, Virginia. In the 1990s, CIA, NSAAnd an RAND corporation computer scientist individually came up with three translations of scrambled letters in four panels of the scalp. However, the final section, known as K4, was encoded with the notatia techniques and it remained unimaginable. This failure has only Obsessed Must be thousands of cryptanalists. When none of them think that they have an answer, they write letters to confirm Jim Sanobar. Sanobar is the artist who created the installation and the only person who knows the answer. The momentum has risen lately. And Sanobar is being sustained – though not for the reason you think.

Consider the email from a recent codebreaker. “It took 35 years and even the NSA I was able to do in just 3 hours before eating my morning coffee,” it started, what they believed to the author Sanobar that they believed to be a cosmic solution. “Rewrite of history,” wrote the submitter. “100% crack in any error.” You may ask, what enables someone to believe that they have surpassed the most elite mathematicians and cryptologists in the world, including some spooks who are likely to Have a quantum computer In the basement? The answer is authentic 2025: a chatbot!

It turns out that the current generation of AI models is happy to accept prompts to solve the cryptos, with the decoded message on the plaintex and announces victory. Sanobar says he is looking at it even more. Of course, this author’s “solution” died was wrong, as was the bounce before thousands of Sanoba.

Sanobar recently contacted me to express his hatred with this development. “It seems like a big shift,” he said. “The number [of submissions] Growing dramatically. And the character of the emails is different – people who have crack their code with AI are perfectly sure they crack the cryptos during the morning! AI seems to be lying to them, each of them told that it is 99.99% sure they are cryptos, congratulations cracks. So they are all sure they have crack it as they reach me. “

It annoys Sanobar in various ways. Until recently there was an unprotected agreement between artists and Cryptos faithful that the attempt to crack the code would be taken seriously. (A few years ago, Sanobar started Charging $ 50 To review the solutions, provides a speed bump for filtering wild estimates and nut cases)) Cryptos’ artistic nature that was fed backwards; In the CIA’s backyard, there is a destructive commentary on the Fanhaus-Miro aspect of an intelligence collection that deny the solution, where every truth is thrown into doubt. This is true that thousands of people have spent a lot of effort to unveil plaintexts – from which judging from The decidated panels so far, Sanobar’s message indicates a shiny on privacy. Newcomers seem to have no idea of ​​this complexity.

Sanobar says, “People who try to crypts cryptos have no idea what the crowd is.” He uses himself to drive himself through Randos’ email using AI shortcuts, which requires very little thought and skills, can be appreciated for the challenge. It is worth saying that you have made a helicopter on the top and scales Everest-but worse, because these ankles-bitter have not solved the code at all. They have just gone up the sea level. Sometimes, in his answers, Sanoba is not behind. He told a stray guesser, “I guess from your certainty that you have used AI.” “AI is false, and not enough information.”