In the Inverted World, the government does not stand at a podium debunking ESP. It quietly cuts a budget line, recruits a handful of people who claim they can mentally project themselves across the planet, points them at Soviet weapons sites, and keeps the operation running for twenty years. This isn't a thought experiment. It is the documented history of the program eventually code-named Stargate, and you can read the files yourself, because the CIA put them online.

The basic facts are not contested by anyone, believer or skeptic. Beginning in the early 1970s, in the thick of the Cold War, the U.S. defense and intelligence establishment funded research into "remote viewing" — the claimed ability to perceive a distant or hidden target using only the mind. The work ran under a rotating series of cryptonyms — Gondola Wish, Grill Flame, Center Lane, Sun Streak, and finally Stargate — across Army Intelligence, the Defense Intelligence Agency, and the CIA, with key research conducted at the Stanford Research Institute by physicists Russell Targ and Harold Puthoff. The motive was straightforward Cold War paranoia: U.S. agencies believed the Soviets were pouring money into "psychotronics," and could not afford to be caught flat-footed if it worked.

The extraordinary part is that this is all in the public record now. In 2017 the CIA released its CREST archive, and the Stargate collection — tens of thousands of pages of session transcripts, protocols, and assessments — sits on the agency's FOIA Electronic Reading Room for anyone to download. The files include the operational sessions of remote viewers like Joseph McMoneagle, the Army's celebrated "Remote Viewer No. 1," and they are not coy about what the program attempted: viewers were tasked against Soviet submarine construction, hostage locations, and hidden facilities, sketching what they claimed to perceive. Some sessions, in the documents, are logged as strikingly accurate; the project's defenders point to a viewer's reported description of a massive new Soviet submarine under construction in a building away from the water.

Now the evidence has to be read honestly, and the strongest evidence in this story actually cuts against the psychics. When the program was shut down, the CIA commissioned an outside evaluation from the American Institutes for Research. That 1995 report — also public — concluded that remote viewing had never been shown to produce intelligence of value in actual operations, that apparent hits were vague enough to be retrofitted to almost anything, and that the laboratory statistics, even where above chance, were too small and too fragile to support real-world tasking. The CIA used that report as the basis to cancel and declassify Stargate. The government's own final verdict was thumbs-down.

So here the skeptical reading and the documentary record point the same way, and that is more interesting than a simple debunk. The remarkable claim is not that remote viewing worked — the official evaluation says it didn't, reliably. The remarkable, fully-proven claim is that the United States government spent roughly two decades and millions of dollars seriously testing whether the human mind could function as a sensor, treated the question as a live national-security matter rather than a joke, and generated a paper trail it eventually handed to the public. That a serious intelligence apparatus took psi seriously enough to fund and field it is itself the anomaly, regardless of whether the talent was real.

And the door does not fully shut. Even the skeptical AIR review acknowledged that some laboratory results were statistically above chance and could not be cleanly explained away — they were simply judged too weak and too unreliable to act on. Believers read that as a faint signal the government lacked the patience to develop; skeptics read it as noise dressed up by selection bias. The files are right there, declassified, and they let you argue either side with primary sources in hand.

That is the unresolved question Stargate leaves on the table. We know the program existed, who ran it, what it was aimed at, and that its sponsors ultimately walked away unconvinced. What we don't have is a clean answer to why an organization as ruthlessly results-driven as the CIA kept the thing alive for twenty years before pulling the plug — and whether the residual, unexplained statistical blips in the lab data are the dying static of a bad idea, or a signal nobody had the nerve to chase.