For 35 years, amateur and professional cryptographers have tried to crack the code on Kryptos, a majestic sculpture behind the CIA headquarters in Lane, Virginia. In the 1990s, the CIA, NSA and a Rand Corporation Computer Scientist came independently of translations for three of the four panels of the sculpture. But the last segment, known as K4, was coded with knotty techniques and remains unsolved. This failure has only deepened the obsession of thousands of potential cryptanalytsen. When one of them thinks they have an answer, they write Jim Sanborn for confirmation. Sanborn is the artist who made the installation and the only person who knows the answer. The pace has been picked up lately. And Sanborn is checked – although not for the reasons you might think.
Consider the e-mail from a recent potential code breaker. “What lasted 35 years and even the NSA with all their resources could not do it that I could do in just 3 hours before I even had my morning coffee,” it started before the writer Sanborn showed what they thought she was the cosmic elusive solution. “History rewritten,” wrote the submitter. “No mistakes 100% cracked.” You may ask, what does someone enable to believe that he had the most elite mathematicians and cryptologists in the world, including some ghosts who might have a quantum computer in the basement? The answer is Puur 2025: a chatbot!
It appears that the current generation of AI models is happy to accept instructions that are aimed at solving Kryptos, coming up with the decoded message in flat text and explaining the victory. Sanborn says he sees it more and more. Of course, the “solution” of this writer was completely wrong, as the thousands of Sanborn had previously bounced.
Sanborn recently contacted me to express his disgust about this development. “It feels like a big shift,” he says. “The figures [of submissions] have increased dramatically. And the character of the E -mails is different – the people who have done their code with AI are completely convinced that they cracked Kryptos during breakfast! AI seems to be lying to them and tells them all that it is 99.99% certain that they have squatted Kryptos, congratulations. So they are all convinced that by the time they reach me, they cracked it. '
This disturbs Sanborn in different ways. Until recently, there was an unspoken agreement between the artist and the Kryptos believers that the effort to crack the code would be taken seriously. (A few years ago, Sanborn started asking $ 50 to assess solutions, offering a speed bump to filter out wild guesses and nuts.) That was introduced back and forth in the artistic nature of Kryptos; Having an object that defies solution in the backyard of the CIA is a subversive commentary on the Funhouse-Mirror aspect of collecting information, whereby every truth is questioned. The fact that thousands of people have spent an enormous amount of effort to reveal the flat text – which, based on the decoded panels so far, indicates that Sanborn's message is a shine on the secret itself. Newcomers seem to have no idea of this complexity.
“The crowd of people who try to crack Kryptos today has no idea what Kryptos is,” says Sanborn. He notices that he is used by e -mails of Rando's using AI fast couplings that require little thinking and expertise, let alone appreciation for the challenge. It is as if you say that you have scaled Everest by making a helicopter ride to the top but even worse, because these single biters have not solved the code at all. They have hardly climbed above sea level. Sometimes Sanborn does not entail his answers. “I concluded from your certainty that you used AI,” he told a miserable Gisser. “Ai lies and does not have enough info.”