Elon Musk announces xAI

Background

Elon Musk had co-founded OpenAI in December 2015 and left its board in 2018.^[2] xAI therefore came after his earlier involvement with that organisation. In July 2023 he made his next AI company public.

What happened

On 12 July 2023 Elon Musk publicly announced xAI. He had incorporated the company in Nevada on 9 March 2023, so the announcement came a little over four months after the legal founding.^[1] The founding team of eleven was led by chief engineer Igor Babuschkin and included Jimmy Ba, Zihang Dai, Kyle Kosic, Manuel Kroiss, Ross Nordeen, Toby Pohlen, Christian Szegedy, Yuhuai Wu, Greg Yang and Guodong Zhang.^[1] The chief engineer is named first, as in the sources, and the team was small enough to be listed in full at the time of the announcement.^[1]

Key details

Musk said xAI would build a “maximally truth-seeking” AI.^[1] That phrase is a statement of intent rather than a measurable standard, and it is recorded here as Musk stated it. A large language model, the kind of system xAI went on to build, is trained on very large amounts of text so that it can understand and generate language, and the company’s goal concerned the answers such systems give.

Less than four months after the announcement, on 3 November 2023, xAI announced its first model, Grok-1, following a 33-billion-parameter prototype called Grok-0.^[3] In November 2023 xAI raised $135 million in equity at a post-money valuation of $673 million.^[1]

Impact

The announcement made xAI public. Its first model and chatbot arrived within four months, and the funding round in November 2023 followed the announcement.^[1][3]

Later events in the company’s history, including its funding rounds, its first model release and its ownership changes, all follow from this starting point. The details recorded here, including the incorporation date, the founding team and the stated goal, are the fixed reference points for that history.^[1]