HOW A QUACK DOC'S MAD QUEST FOR TELEPATHIC POWERS LEAD TO MODERN AI?

All hail AI: the missing link in medical innovation and drug discovery, the long-heralded tool for human salvation, and the technology that was promised by the philosophers of yore.

Forgive my theatrics, it does seem fitting for an article about modern AI though.

After all, AI tools have taken hold of innovation in health and science in many inspirational – and unsettling – ways. Nearly every major scientific journal has published papers on the use of AI in their specific field, from benchmarking studies to editorial ethics statements.

The truth is I don’t believe that AI is either a savior or future enslaver of humanity. As a health and science writer who has spent years writing about its pragmatic uses and flaws in those fields, I find I’m far more interested in how oddly human AI seems to be.

To that point, I would like to share my favorite example of some weird humanity at the center of this technology:

You’ve probably heard that AI models use artificial neural networks, like those in our brains, to compute vast amounts of information. But do you know how those neural networks – the foundational concept of AI technology – were developed in the first place?

It all started with a “quack” German physician by the name of Hans Berger, whose unusual contributions to medicine and the study of the human brain became surprisingly useful in the world of advanced computing.

Berger, who was born in the small German town of Neuses on May 21, 1873, was a textbook example of a medical gadfly. As a kid, he dreamed of becoming an astronomer and mathematician. Unfortunately, he failed to reach those academic goals, so instead of equations and stargazing, he joined the German cavalry when he was 19 years old.

But his fortunes did not improve. During a training exercise, Berger fell off his horse onto the ground directly in front of another horse who nearly trampled him to death. Berger lived, but this terrifying near-death experience would be a major turning point in his life – and potentially all our lives. Just not in the way you might expect.

After his fall and before he could write home, Berger received a letter from his father telling him that his sister had a sudden sense of dread that he had died, which might seem normal except that she was many miles away from Berger at the time and had no knowledge of his close brush with death-by-horse. Possibly even less normal, however, was Berger’s interpretation of this letter and his sister’s concern: he became convinced that humans possessed telepathy.

So Berger went back to school at the University of Jena to study psychiatry and neurology. In 1897, he received a medical degree and started working at the university’s clinic. This is where he began his experiments on the human brain to unlock its telepathic potential.

While Berger never found any evidence that telepathy was possible, he did make a notable discovery that would shape the future of medicine, science, and computational engineering.

By the early 1920s, Berger was considered by his German contemporaries to be a quack. Despite those doubters, Berger still managed to become an expert – as gadflies often do – in the emerging field of electrical activity in the human brain. Most importantly, he developed a way to measure it.

In his desperate and doomed efforts to discover telepathic power, Berger created the first ever electroencephalogram, more commonly known as an EEG, and this device would prove to be a critical tool in many groundbreaking discoveries, such as the existence of brainwaves.

By the mid-1930s, the EEG shifted from being a scientific oddity to being a critical clinical research tool for a wide range of conditions, including brain tumors, strokes, epileptic seizures, sleeping disorders, and even explorations of consciousness. It also aided in the discovery of billions of interconnected neurons in the brain, which came to be known as neural networks.

Enter Warren S. McCulloch, an American scientist and another gadfly of his time, who made notable contributions to medicine, engineering, and cybernetics, which is the study of self-regulating systems. Sounds familiar, right?

McCulloch was known as “an intellectual showman” who adopted numerous professional identities throughout his life, including philosopher, neurologist, neurophysiologist, neuropsychiatrist, engineer, and – just to be more well-rounded I guess, poet.

Despite the diverse range of interests, he claimed all those roles were in service of a single pursuit: understanding the relationship between the mind and the brain. It’s worth noting that both Berger and McCulloch made incredible contributions to the development of AI, while passionately pursuing answers to completely unrelated problems.

It might also be worth noting that neither man succeeded in finding their answer – the brain-mind connection remains a mystery and we still don’t have telepathic powers. So maybe give yourself a break; you aren’t failing, you’re just setting up humanity for a major technological leap… in the distant future.

Anyway, McCulloch eventually started collaborating with Walter Pitts, a logician, neuroscientist, and a real-life Will Hunting. The two men ended up working on computational models using mathematical algorithms based on something called threshold logic. With McCulloch’s research on brain activity and the nature of neurons and Pitts’ mastery of logic and computing, they started publishing papers that would establish the foundation for current AI models.

Eventually, they published a paper titled “A Logical Calculus of the Ideas Immanent in Nervous Activity,” which introduced for the very first time the idea of artificial neural networks in computing.

In that truly groundbreaking entry into the scientific literature, Pitts and McCulloch explained how artificial neural networks could compute information more effectively and efficiently – and exactly like the human brain. This paper was published in 1943 – roughly 80 years before ChatGPT 3.5 took the internet by storm.

Whether you think AI is a herald or harbinger, you can thank or curse three true gadflies of medical and scientific research. While McCulloch and Pitts were more respected and contributed more directly to this ever-evolving technology, it was the ideas and obsessions of a “quack” German doctor who made a critical, albeit unheralded, contribution to AI.

In a way, the very foundation of AI computing was built on the mad pursuits of an unusual scientist who just wanted to have a telepathic connection with his sister.

For me, knowing this part of its origin story makes AI technology feel a little more human – and a lot weirder. With DNA like that, it might not be so bad after all.

By Michael DePeau-Wilson, Health and Science Contributor

*Photo Credit: Medical Republic