Chapter 133

Musk conceives of Neuralink, a brain-computer interface company founded in late 2016 to implant chips directly into human brains. Drawing inspiration from science fiction and frustrated by the slowness of traditional input methods like typing, he aims to create a symbiotic connection between human minds and machines that could transmit information a million times faster than current interfaces.

Musk traces the history of human-computer interfaces from J.C.R. Licklider's 1960 vision of human-machine symbiosis through video displays and graphical user interfaces to voice-activated systems like Siri. In 2016, while typing on his iPhone, Musk realizes that brain-to-machine communication at roughly 100 bits per second is insufferably slow and asks Sam Teller to find a neuroscientist to explore brain-computer interfaces. He envisions a chip implanted in the skull that could transmit information at speeds a million times faster, inspired by Iain Banks' "neural lace" concept from science fiction. Musk founds Neuralink in late 2016 with cofounders including neuroscientist Max Hodak and engineer DJ Seo, who had independently developed "neural dust" implants at Caltech and Berkeley. He also recruits Shivon Zilis, a tech investor and AI enthusiast who becomes a key manager and close personal companion. The technical foundation rests on the Utah Array, a 1992 microchip with 100 electrodes; Neuralink's 2019 design improves on this with over 3,000 electrodes on 96 threads. In August 2020, Musk presents a quarter-sized chip implanted in a pig named Gertrude that wirelessly transmits brain signals to a computer. Months later, when engineers show Musk a complex four-chip design with multiple wires and routers, he rejects it as over-engineered and demands a single, elegant device with no connections—triggering initial frustration but ultimately driving the team to develop a single-chip solution with Bluetooth transmission. The chip size becomes another point of contention; Musk questions whether the skull size requirement is fundamental or arbitrary, leading to a larger design that proves more feasible. The team successfully implants the improved chip in a macaque named Pager, who learns to play Pong by moving a joystick, then controls the game with brain signals alone—a major proof of concept that generates 6 million YouTube views within a year.