In 2034, the first person landed on Mars. While she didn鈥檛 go there physically, she still experienced the planet intimately. She explored an ancient river delta and built a base. She put up a flag (China鈥檚) and conducted a detailed analysis of rock samples. She achieved all this by inhabiting a robot via a sophisticated brain-computer interface. Some people claimed the woman had 鈥� in a real sense 鈥� been to Mars.
Critics said she hadn鈥檛, because her body was always in a lab in Beijing. Ah, but her mind was on Mars, replied her supporters. That there was even an argument demonstrates how far brain-computer interfaces (BCIs) had improved. And, as with many advances, artificial intelligence proved the key.
The basic idea of a BCI is to pick up the signals of a living brain and enmesh them with the electronics of a computer. It is easier said than done. Brain activity is messy, coming from multiple locations and having overlapping and interfering properties and imprecise meanings. But improvements in neuroengineering and machine learning able to interpret the signal made it possible to record complex brain activity from electrodes planted into a subject鈥檚 cortex, and translate it into a range of actions.
Advertisement
The technology was initially driven by surgeons hoping to improve the quality of life for people with paralysis or locked-in syndrome. The first BCI using electrodes was in a paralysed man in 1998, consisting of just two electrodes. The man became able to spell out words on a computer by moving the cursor with his mind. Such breakthroughs were world-changing for people who were otherwise unable to communicate.
The problem was that electrodes were few, while neurons numbered in the billions. Improvements in micro-electrode construction allowed neuroscientists to record more brain activity; in 2014, Juliano Pinto, a paraplegic unable to move his legs, inhabited an exoskeleton and, using mind control, kicked the first ball at the FIFA World Cup in Brazil.
In 2023, a used an 鈥渆lectrocorticography鈥� array placed on the brain 鈥� but without penetrating it 鈥� to pick up signals from a paralysed man鈥檚 motor cortex and relay them to the spinal cord. An interface at the spinal cord then sent the signals to the man鈥檚 legs, allowing him to walk again. A similar interface allowed a woman unable to speak after a stroke to and speak through it. By 2025, a paralysed man was able to pilot a drone; eventually people could 鈥渋nhabit鈥� such drones.
Consciousness had relocated to the robot's body 鈥� the human felt that they had become the avatar
By the end of the 2020s, paralysed people could inhabit exoskeletons and robots by sending their thoughts to the machine, just as people had done to control cursors on a screen. The difference was that signals also went the other way: from machine to human. The human controller鈥檚 senses were fed directly from the robot鈥檚 cameras and microphones, and haptic feedback 鈥� the transmission of touch using vibration and pressure suits 鈥� helped give the user a 鈥渞eal鈥� sense of location. Retinal and cochlear implant technology had progressed enough that scientists could override the direct input into someone鈥檚 eyes and ears, and replace them with inputs from a remote feed. Similar inputs enabled smell and taste.
The upshot of all this: people could operate a robot remotely and see, hear and feel things that it interacted with. They were able to explore and experience extreme environments 鈥� the crushing depths of the ocean, the caldera of an active volcano. The boundary between self and machine softened, especially when the robot could interpret , building on work by at the Australian National University in Canberra.
Robots would move faster if the human controller was happy and excited, and more cautiously if the person was nervous. A facial display on the robot would indicate a feeling such as happiness or confusion, and then a human could understand what the robot was 鈥渇eeling鈥�. The user鈥檚 consciousness relocated from their body to the robot鈥檚; the human felt they had become the avatar, hence the argument that the telepresent robot really did represent the first human on Mars.
Such was the success of these space missions that it was years before state-owned space agencies committed to the risk and expense of sending humans. Even greater success came with rescue robots, as they were called, which were used as avatars to enter burning buildings or war zones and evacuate people. The rescue robots name, however, came to refer to the escape that paralysed and locked-in people were able to achieve from their beds.
Rowan Hooper is 快猫短视频鈥榮 podcast editor and the author of How to Spend a Trillion Dollars: The 10 global problems we can actually fix. Follow him on Bluesky @rowhoop.bsky.social
听
