Telepathic communication is the transfer of information from one mind to another without using speech, writing, gestures, or any of the ordinary senses. The word itself comes from the Greek “tele” (distant) and “pathe” (feeling or occurrence), and it was coined in 1882 by the historian Frederic Myers. For most of its history, telepathy has been a concept rooted in the paranormal. But in the last decade, neuroscience and engineering have started building real, technology-assisted versions of mind-to-mind communication, blurring the line between science fiction and laboratory fact.
The Traditional Idea of Telepathy
Telepathy as most people understand it is a form of extrasensory perception, or ESP. The idea is that one person can send thoughts, images, or feelings directly into another person’s mind without any physical medium. This concept has appeared across cultures for centuries, but it became a serious topic of investigation in the late 19th century, when researchers began trying to test it under controlled conditions.
The first sustained laboratory effort started in 1927, when J.B. Rhine, a botanist turned psychologist at Duke University, launched a series of card-guessing experiments that essentially founded the field of parapsychology. Over the following decades, researchers developed more sophisticated tests. The most well-known modern protocol is the Ganzfeld experiment, in which a “receiver” sits in a state of sensory deprivation while a “sender” in another room tries to transmit a randomly selected image. Meta-analyses of these experiments have reported statistically significant results, but with important caveats.
A 2016 analysis published in PLoS ONE examined the Ganzfeld database and found that questionable research practices, such as selective reporting and flexible stopping rules, could account for roughly 60% of the reported effect size. The overall result remained technically significant even after correction (p = 0.003), but the authors noted that the phenomenon being investigated “is widely believed to be nonexistent” by the broader scientific community. The prevailing view among physicists and neuroscientists is that no known mechanism exists for biological telepathy. Neurons operate at scales where quantum effects, sometimes proposed as a possible explanation, are generally considered irrelevant to brain function.
Technology That Mimics Telepathy
While unaided mind-to-mind communication remains unproven, researchers have built working systems that accomplish something functionally similar using technology as a bridge. These systems are called brain-to-brain interfaces, and they work in two stages: reading useful information from one person’s brain activity, then writing that information into another person’s brain using stimulation.
In a landmark experiment first demonstrated in August 2013 at the University of Washington, researchers connected two people sitting in separate buildings on campus. The “sender” wore an EEG cap that recorded electrical activity from their scalp. When the sender imagined moving their hand, a computer detected that intention, transmitted it over the internet, and delivered a magnetic pulse to the “receiver’s” motor cortex using transcranial magnetic stimulation (TMS). The pulse caused the receiver’s hand to physically move and press a touchpad, completing a cooperative task that neither person could have done alone.
A 2014 experiment took this a step further across continents. A sender in France thought of the word “ciao.” An EEG picked up the brain signal, a computer translated it into binary code, and the data was sent over the internet to a receiver in India. The message was converted back into small pulses delivered to the receiver’s brain. Brain imaging revealed that a specific memory-related brain region activated in the receiver during successful transmission but not in control subjects. These experiments are slow, low-bandwidth, and require elaborate equipment, but they represent genuine information transfer from one brain to another.
How Brain-Computer Interfaces Work
The technology behind these experiments relies on brain-computer interfaces (BCIs), which translate neural activity into digital commands. On the reading side, researchers can pick up brain signals using methods that range from non-invasive (EEG caps on the scalp) to highly invasive (electrodes implanted directly in brain tissue). Implanted electrodes capture far more detail but require surgery. EEG is safe and portable but picks up a noisy, blurred version of what neurons are doing.
On the writing side, information can be pushed back into the brain through magnetic pulses (TMS), focused ultrasound, or direct electrical stimulation through implanted electrodes. Each method has tradeoffs in precision, safety, and how much information it can deliver per second.
The biggest technical hurdle is signal quality. EEG signals have an inherently low signal-to-noise ratio, meaning the actual brain activity you want to detect is often buried under electrical noise from eye blinks, muscle movements, and other artifacts. Classifying something as complex as imagined speech from EEG data remains, as one IEEE study put it, “significantly challenging” because muscle signals from the face and scalp can dominate and obscure the cortical signals researchers are trying to read.
Real-World Applications for Paralyzed Patients
The most immediate and impactful use of this technology isn’t sending thoughts between two healthy people. It’s restoring communication for people who have lost the ability to speak or move. Patients with locked-in syndrome, often caused by brainstem strokes, are fully conscious but unable to control any muscles except sometimes their eyes. For these individuals, a brain-computer interface can be a lifeline.
In one study, a locked-in patient used an EEG-based system called a P300 speller to communicate over 62 sessions spanning more than a year. The system flashed letters on a screen and detected which letter the patient was focusing on by reading a specific brainwave response. A standard keyboard layout proved too complex at first, yielding only about 32% accuracy. But when researchers simplified the interface to four choices (yes, no, pass, end), accuracy jumped to an average of 94.7% across all sessions. That simple shift turned an unreliable system into a functional communication tool.
A related but distinct technology uses sensors on the face and neck to detect the tiny muscle signals people produce when they mouth words silently. These “silent speech interfaces” can convert subvocal speech into text or synthesized audio without any sound being produced. Unlike direct brain reading, this approach picks up muscle activity rather than neural signals, making it simpler and more reliable with current technology. It’s particularly useful for people who have lost their voice box but retain control of their facial and neck muscles.
Neuralink and the Next Generation
The most high-profile effort to push brain-computer interfaces forward is Neuralink, which has developed a coin-sized implant designed to be “cosmetically invisible” once placed beneath the skull. The device uses 1,024 electrodes woven into the brain’s outer layer by a surgical robot, targeting individual neurons rather than reading broad electrical patterns from the scalp. Neural signals are processed by a custom chip inside the implant and transmitted wirelessly via Bluetooth to external devices.
The company’s initial focus is on people with severe paralysis, aiming to let them control phones, computers, and prosthetic limbs through thought alone. The longer-term vision includes treating neurological conditions like Parkinson’s disease and, eventually, enabling direct thought-to-device communication for a broader population. Neuralink has named its first product “Telepathy,” a deliberate nod to the concept, though what it actually does is closer to a very precise brain-to-computer link than mind-to-mind contact.
The gap between where the technology is now and anything resembling the telepathy of popular imagination remains enormous. Current systems can transmit simple commands: move a hand, select a letter, answer yes or no. Sending a complex thought, a visual memory, or an emotional state from one brain to another in real time would require reading and writing neural activity at a resolution and speed that no existing technology approaches. But the direction of progress is clear, and each generation of hardware closes the gap a little further.