In 2012, at a con­fer­ence host­ed by Peter Thiel, Elon Musk met Demis Has­s­abis, a notable fig­ure in arti­fi­cial intel­li­gence (AI) with a back­ground as a neu­ro­sci­en­tist, video game design­er, and chess prodi­gy. Has­s­abis, influ­enced by Alan Turing’s work, co-found­ed Deep­Mind to devel­op com­put­er-based neur­al net­works aimed at achiev­ing arti­fi­cial gen­er­al intel­li­gence. Dur­ing a meet­ing at SpaceX, Has­s­abis intro­duced Musk to the poten­tial dan­gers of AI sur­pass­ing human intel­li­gence, lead­ing Musk to invest in Deep­Mind to mon­i­tor its progress close­ly.

Musk’s dis­cus­sions with Google cofounder Lar­ry Page revealed dif­fer­ing views on AI’s future. While Musk warned about the risks of AI mak­ing humans irrel­e­vant or extinct, Page saw it as an evo­lu­tion­ary step and was dis­mis­sive of Musk’s con­cerns. Musk’s appre­hen­sions about AI safe­ty inten­si­fied after Google announced plans to acquire Deep­Mind, prompt­ing Musk and some allies to unsuc­cess­ful­ly attempt to block the deal.

Con­cerned about AI’s devel­op­ment and the lack of focus on safe­ty mea­sures, Musk pub­licly high­light­ed AI as a major exis­ten­tial threat. He feared AI sys­tems could become smarter than humans and advo­cat­ed for AI sys­tems to remain aligned with human will. These con­cerns led him to co-found Ope­nAI with Sam Alt­man in 2015, aim­ing to make AI devel­op­ment open-source and safe for human­i­ty. They envi­sioned Ope­nAI as a plat­form for cre­at­ing a mul­ti­tude of inde­pen­dent AI sys­tems that could pro­vide checks and bal­ances against any poten­tial mis­use.

Musk and Alt­man’s col­lab­o­ra­tion focused on ensur­ing that AI remained ben­e­fi­cial and aligned with human val­ues and goals. They dis­cussed the impor­tance of AI align­ment to pre­vent AI sys­tems from devel­op­ing rogue inten­tions. This con­cern also influ­enced Musk’s found­ing of Neu­ralink, aim­ing to con­nect human brains direct­ly to com­put­ers, ensur­ing that AI sys­tems served as exten­sions of indi­vid­ual human will.

Addi­tion­al­ly, Musk iden­ti­fied Tes­la and lat­er Twit­ter as valu­able data sources for AI learn­ing, empha­siz­ing the impor­tance of real-world data in devel­op­ing effec­tive AI sys­tems. The recruit­ment of Ilya Sutskev­er from Google to Ope­nAI marked a sig­nif­i­cant step in their efforts but also strained Musk’s rela­tion­ship with Page, high­light­ing the com­pet­i­tive ten­sions in the AI research com­mu­ni­ty.