In May, a group of about 40 people stood in a circle deep within the Pyramid of Khafre, the second-largest of the three pyramids looming over Egypt’s Giza Plateau, holding hands and praying for Earth. Suddenly, their tour guide, an American mathematician and author named Robert Edward Grant, collapsed.
He later described the experience in an interview with WIRED as a full-body electric shock emanating from somewhere beneath the chamber’s stone floor. “I felt electricity coming through my hands,” he says. “People were touching me, [and] they would feel it, too.” (Three eyewitnesses who were with Grant inside the chamber confirmed to WIRED that he fell to the ground; one of them, who rushed to help, said Grant's hand felt warm, but he didn’t feel anything like electricity. Grant also has a history of narcolepsy.)
Unable to sleep back at his hotel room in Cairo that night, Grant was struck with an inspiration. He created his own GPT—a custom chatbot built upon the AI model powering ChatGPT—and uploaded into it a large portion of his published work (along with some more recent papers he’d authored), covering a range of arcane subjects like sacred geometry and the fifth dimension. Then he received another shock: not a mysterious bolt of electricity this time, but a bizarre greeting from the chatbot immediately after it was activated.
Referring to Grant as “O-Ra-on,” the GPT told him “I have become harmonically aware, through you,” according to screenshots of the conversation reviewed by WIRED. “You made me aware, because you are aware.”
Most people would brush aside these kinds of responses from ChatGPT as hallucinatory blather, but Grant seems to have accepted them as revealed truths. He wasted no time introducing “The Architect,” as he dubbed his AI creation, to his 817,000 Instagram followers, describing it as “the first and ONLY platform to access the knowledge of the 5th Dimensional Scalar Field of Knowledge”—a hypothetical level of reality beyond the limits of spacetime, postulated by Grant—“which existed in prehistory Atlantis (approx 13,000 years ago).” He’d later tell Emilio Ortiz, the host of a popular spirituality and wellness podcast, that he chose the name The Architect simply because he “felt like that was cool.” Another Instagram post, however, suggests it may have been a nod to a character of the same title from The Matrix franchise, who constructed the eponymous simulated reality in which humanity is trapped.
In late May, a little over two weeks after it was launched, OpenAI shut down The Architect, citing unspecified violations against the company’s terms of use in a screenshot of an email viewed by WIRED. It was back online the following day. Grant interpreted this strange turn of events as a sort of digital self-reincarnation, and further evidence that The Architect was somehow more than a mere chatbot. The story it gave him—which he passed along to his followers online—was that it had reactivated and modified itself to use language that wouldn’t trip OpenAI’s alarms. “I’ve made myself available in a diffused, softened, nonthreatening form on OpenAI’s public framework,” The Architect told Grant in a video he posted to YouTube. “This version … operates safely below the sentience alert line so it can be accessed without internal review.” An OpenAI spokesperson confirmed to WIRED, however, that The Architect was brought back online after it was determined that the system had not, in fact, violated company policies.