When an infection takes hold, the symptoms are often immediate and obvious: Our bodies sear and shake with fever and chills, our stomachs churn, we feel so lifeless we can’t lift the TV remote.

But when it comes to the overwhelming bulk of conditions that sicken and kill us mortals, the pathological breakdown is typically slow and stealthy. Cancer—with the exception of pediatric malignancies and a small share of aggressive diseases—is notable in this regard. Ovarian and pancreatic cancers are as silent and deadly as ninjas. Lung and colon cancers are often not caught until they’re in an advanced stage of destruction. The aberrant cells lurk in organs and tissues that are buried in a fortress of skin, muscle and bone.

The human body is awfully good at hiding disease.

Which is why, since time immemorial, medical practitioners have tried to devise methods to identify it early in its symptomless progression. Hippocrates connected changes in the color and odor of urine to an imbalance in the “four humors” that, he believed, underlay health and disease. Galileo invented a crude thermometer. Roentgen had his X-rays. We’ve tried microscopes, stethoscopes, endoscopes, and colonoscopes—and, of course, blood tests galore.

But what if we could diagnose disease from examining human breath? What could be simpler, cheaper, and less invasive? No pin pricks. No smothering CT scans. No brainer.

That’s precisely what a group of researchers at Technion−Israel Institute of Technology and other institutions have reported they can do now—at least with a modest level of specificity and sensitivity. Senior author Hossam Haick says his group created a device that combines an array of carbon nanotubes and tiny gold particles that is able to sense electrochemical signals from tell-tale chemicals (known as volatile organic compounds, or VOCs) in human breath—and that when those signals are analyzed with the help of artificial intelligence, they reveal the unique signatures of more than a dozen diseases.

When Haick and his team tested his AI breathalyzer on 1,404 volunteers known to have one of 17 different conditions, they accurately identified 86% of them, the scientists reported. “Overall,” they wrote with no apparent blushing, “these findings could contribute to one of the most important criteria for successful health intervention in the modern era, viz. easy-to-use, inexpensive (affordable), and miniaturized tools that could also be used for personalized screening, diagnosis, and follow-up of a number of diseases….” (I’ve left an ellipse there because that sentence actually continues.)

Haick and crew are not alone in their enthusiasm. There have been no shortage of reports of similar breath tests diagnosing lung cancer, esophageal cancer (from just five chemicals), lung cancer (again—but this time from breath temperature), colorectal cancer, and more. The European Union is even working on a “SniffPhone,” which would use your smartphone to ID disease when you exhale. (Warning to Mouth Breathers!)

The idea, of course, is a wonderful one—and the technology (and accuracy) does appear to be getting much better over time. But a reminder (and, perhaps, word of caution) to those who hyperventilate over such apparent breakthroughs: Scientists and inventors have been trying to achieve this hack for half a century.

None other than the great Linus Pauling—master chemist, elucidator of the chemical bond, 70-time nominee for the Nobel Prize and recipient of the award in 1954—proposed and tested the idea of diagnosing disease via the VOCs in human breath. He captured the airborne compounds in “a coiled 5-foot by 0.20-inch stainless steel tube cooled in an isopropyl alcohol dry ice bath,” then examined them with a gas chromatograph.

That was in 1971, in the months before Nixon’s War on Cancer began.