Elon Musk, the CEO of Twitter, has announced the launch of a new organization called xAI. The purpose of xAI is to delve into understanding the true nature of the universe. Although this may be seen as ambitious, Musk plans to provide more details during a Twitter Spaces event on Friday.
The team at xAI currently consists of 12 individuals and is led by Musk himself, along with veterans from reputable organizations such as DeepMind, OpenAI, Google Research, Microsoft Research, Tesla, and the University of Toronto. As for advisory support, xAI will be guided by Dan Hendrycks, the director at the Center for AI Safety - an AI research nonprofit. Additionally, xAI will collaborate with both Twitter and Musk’s other company, Tesla, in order to make progress towards its mission.
Even though specific details are being saved for Friday’s event, some insights can be gleaned from xAI’s roster. Many founding members possess expertise in large language models similar to OpenAI’s GPT-4 or have experience in reinforcement learning techniques. Reinforcement learning involves training AI models to perform tasks and rewarding them accordingly. Musk expressed his intention to create a “maximum-truth-seeking AI” called “TruthGPT” during an interview with Tucker Carlson in April. It is likely that xAI’s initial efforts will revolve around developing a text-generating AI that Musk perceives as more truthful than existing models. Musk has previously proposed using Twitter data to train ChatGPT-like technology.
Rumors about Musk establishing an AI company have circulated for some time now. Earlier this year, Business Insider reported that Musk had purchased thousands of GPUs to power an upcoming generative AI product. The Financial Times also mentioned that he planned to establish an AI firm as competition for Microsoft-backed OpenAI. Allegedly seeking funding from SpaceX and Tesla investors to kickstart the venture, it seems that Musk’s ambitions in the AI field have grown since his departure from OpenAI.
Musk’s relationship with OpenAI changed significantly after he split with co-founders Sam Altman and Ilya Sutskever a few years ago. Initially launched as a nonprofit organization in 2015, OpenAI received $1 billion in donations from Musk and other contributors. However, as OpenAI shifted its focus towards commercial projects rather than open-source research, Musk became disillusioned and started to view the company as a competitor. He recruited key employees from OpenAI for Tesla’s Autopilot, their driver assistance technology. Furthermore, Musk openly criticized OpenAI, referring to it as a “profit-maximizing demon from hell.” In 2018, he resigned from the OpenAI board and more recently cut off the company’s access to Twitter data due to perceived inadequate compensation for their licensing deal.
In summary, Elon Musk has announced the formation of xAI, an organization aimed at exploring the true nature of the universe. With an experienced team comprised of individuals from prestigious institutions, xAI is set to collaborate with Twitter and Tesla. While specific details are forthcoming, it is expected that xAI will focus on developing advanced language models and truth-seeking AI technologies. This endeavor comes after a shift in Musk’s relationship with OpenAI and his growing ambitions within the AI field.