Are Humans Already Cyborgs, and Have We Always Been?

On October 8, 1958, a Swedish engineer named Rune Elmqvist and a surgeon named Åke Senning implanted the first fully internal cardiac pacemaker into a 43 year old patient named Arne Larsson at Karolinska Institute in Stockholm. The device failed within hours. They implanted a second one. Larsson went on to outlive both men who built and installed it. He died in 2001 at age 86, having received more than twenty replacement units over four decades, his heart running on a rotation of batteries and circuits that nobody in that operating room in 1958 expected to still be ticking in the twenty first century.

I keep coming back to Larsson because his story already answers the question everyone treats as futuristic. He wasn’t part human, part machine in some speculative sense. He was, for the last 43 years of his life, a person whose heartbeat was scheduled by a device his body did not grow. Nobody in 1958 called him a cyborg. Nobody today calls pacemaker patients cyborgs either, even though there are more than a million cardiac devices implanted worldwide each year. We reserve that word for something stranger, something with a whiff of science fiction. That reservation is the mistake.

The Pacemaker Was Never the Beginning

Start with what a pacemaker actually does. It doesn’t add a new capability to the body. It restores a rhythm the heart’s own electrical system, the sinoatrial node, failed to keep. That’s the frame most people use to decide whether a device counts as invasive or radical: does it fix something broken, or does it add something new? Under that frame, pacemakers, insulin pumps, and cochlear implants get a pass. They’re medicine. Brain implants that read intention, or someday write memory, get filed under a different, scarier category.

But the frame doesn’t hold up historically. William House built the first cochlear implant in Los Angeles in 1961, giving deaf patients a crude sense of sound through direct electrical stimulation of the auditory nerve, bypassing the cochlea’s own hair cells entirely. The FDA approved the first cochlear implant for adults in 1984. That device didn’t repair the ear. It replaced the ear’s signaling function with a wire and a processor, permanently, for the rest of a patient’s life. Alim Louis Benabid and Pierre Pollak in Grenoble started implanting electrodes into the thalamus in 1987 to stop tremor by overriding a misfiring circuit with a competing electrical signal, not fixing the misfire, just drowning it out. The FDA cleared deep brain stimulation for essential tremor in 1997 and for Parkinson’s disease in 2002. None of these devices are exotic anymore. They’re standard of care. And every one of them works by the same principle: intercept the body’s electrical signaling and replace or override it with an engineered one.

Your Cells Ran on Voltage Before Anyone Built a Device

Here’s the part that gets lost in every debate about implants and augmentation. The body was never a purely chemical machine that electricity got bolted onto. Every cell in a human body, not just neurons, maintains a voltage gradient across its membrane, typically somewhere around negative 70 millivolts inside relative to outside, held there by ion pumps working constantly against the cell’s own leakiness. That gradient is not a side effect of biology. It is a piece of the code.

Michael Levin, who runs the Allen Discovery Center for Biological Electricity at Tufts University, has spent two decades showing how much of that code we’ve ignored. His lab has demonstrated that bioelectric signaling patterns, voltage gradients across groups of cells, direct where a limb regenerates on a flatworm, what shape an eye takes on a tadpole, whether tissue reads itself as head or tail during development. Genetics builds the parts. Bioelectric patterning tells the parts where to go. Levin’s planaria experiments can produce a worm with a head grafted where a tail should be, purely by altering the voltage pattern across the tissue, no gene editing required.

The body was electric long before anyone plugged it in.

That reframes the entire cyborg question. If cellular identity, organ formation, and neural computation were already running on voltage before a single engineer touched a human body, then the arrival of pacemakers and cochlear implants and deep brain stimulators wasn’t a rupture between biology and technology. It was one electrical system meeting another electrical system it happened to share a vocabulary with. The wire didn’t introduce electricity into the body. It found electricity already there and started talking to it in its own language.

The Interface Era Is an Acceleration, Not a Break

What’s changed in the last fifteen years isn’t the principle, it’s the resolution and the reach. Kevin Tracey’s work at the Feinstein Institutes on the inflammatory reflex showed that a single nerve, the vagus, could be dialed up or down to control systemic inflammation, which is a long way from just pacing a heartbeat or damping a tremor. Vagus nerve stimulation got FDA approval for epilepsy in 1997 and for treatment resistant depression in 2005, and it’s now being tested for rheumatoid arthritis and inflammatory bowel disease, conditions nobody would have described as electrical thirty years ago.

Then the interfaces moved into the brain itself, and the traffic started running both directions. Krishna Shenoy’s work at Stanford, through the BrainGate consortium, used intracortical electrode arrays to let paralyzed patients move a cursor or type by decoding intended hand movements straight from motor cortex, no muscle involved at all. Tom Oxley’s company Synchron took a different route: a stent-like electrode, the Stentrode, threaded through the jugular vein into a blood vessel next to the motor cortex, avoiding open brain surgery entirely. The first human implant happened in Melbourne in 2019. Neuralink implanted its first human patient, Noland Arbaugh, in January 2024, with a fully implanted wireless array reading cortical activity to control a computer.

Each generation of device does the same basic thing the pacemaker did in 1958: intercept an electrical signal the body already generates and reroute it. What’s accelerating is precision and bandwidth, not the underlying premise. We didn’t cross into cyborg territory with Neuralink. We’ve been walking that road since Stockholm.

The Line That Actually Matters Is Agency, Not Anatomy

So if the boundary between body and machine was never as sharp as we assumed, what should we actually be arguing about? Not whether a device is inside the skull or outside it. The question that matters is who holds the control loop, and what happens when the device does more than restore something lost.

Grégoire Courtine and Jocelyne Bloch, working between EPFL and Lausanne University Hospital, published a study in 2018 in which epidural electrical stimulation, timed to a patient’s own intended movement, let three people with spinal cord injuries regain voluntary control of their legs, including a patient named David Mzee who walked without support during the trial. In 2023 the same team published a further step: a “digital bridge,” a brain-computer interface reading intention from the cortex and wirelessly relaying it to spinal stimulators, that let a patient named Gert-Jan Oskam walk again years after a paralyzing accident. That’s not override anymore. That’s collaboration between the patient’s own motor intention and an external circuit built to carry the signal the injury interrupted.

The real question was never whether the wire is there. It’s who holds the other end.

That’s the fork worth arguing about. A pacemaker’s control loop belongs to a cardiologist reading a stored log at a checkup. A Stentrode or a Neuralink implant’s control loop involves a private company’s firmware, cloud infrastructure, and update cycles, sitting between a patient’s intention and their own hand on a cursor. Kevin Tracey’s bioelectric medicine sits closer to the pacemaker model: a clinician dialing a known reflex. The BCI companies sit somewhere newer, where the company that owns the decoding algorithm has more day to day power over the interface than the patient wearing it. Nobody built an equivalent framework for that when insulin pumps and pacemakers were the only implants in town, because none of those devices had ambitions to reshape cognition, mood, or memory. The stentrodes and cortical arrays now in trials do, eventually, carry those ambitions. That’s the genuine break from Arne Larsson’s story, not the presence of the wire, but the presence of an outside party with an interest in what the wire eventually does.

Restoration Was Always the Easy Case

Every device I’ve named so far was built to restore a function biology already had and lost: a heartbeat, hearing, a steady hand, a working leg. That’s the easy ethical case, and it’s the one the FDA has forty years of precedent handling. The harder case, the one the current wave of neural interface companies is explicit about wanting, is augmentation: giving a healthy brain a channel it never had, faster typing, direct machine control, eventually something like an expanded working memory. Restoration borrows legitimacy from medicine. Augmentation has to earn its own, and right now nobody, not the FDA, not the companies, not the bioethicists, has a settled answer for what earning it looks like.

I don’t think the answer is to slow down the hardware. The hardware is going to keep improving because the underlying biology rewards it: electrical signaling is fast, specific, and, as Levin’s work suggests, deeper in the machinery of life than we assumed even at the cellular level. What needs to catch up is the governance of the loop, not the wire itself. Arne Larsson never had to negotiate a terms of service agreement with the company that kept his heart beating. The next generation of implant patients will, and that gap, not the presence of electrodes in a human body, is the real cyborg question of this decade.

Credit: Dave Rodgers on Unsplash