Elon Musk Launches xAI To 'Understand Reality'

Elon Musk has started a new company called "xAI," with Musk and his new venture stating it aims to "understand reality." The company's website, on the other hand, mentions that it seeks to understand the universe, specifically.

Musk recently pitched the idea of creating "a maximum truth-seeking AI that tries to understand the nature of the universe." He proposed the name TruthGPT for it, a subtle jab at ChatGPT and its accuracy woes. It looks like xAI is very much the culmination of those ambitions. The xAI company is not part of X Corp, which is now the parent company for Twitter. However, Musk's latest AI venture will work closely with Twitter, Tesla, and other companies as well. 

Right now, we don't know exactly what the company is planning to build, but the team at xAI will be available to answer all the burning questions in a Twitter Spaces session on July 14. This won't be Musk's first rodeo with AI. Notably, Musk was among the founding members of ChatGPT-maker OpenAI, and also had a seat on its board. However, he reportedly wanted to acquire the company, but his bid was rejected and made a bitter exit.

Will Musk do AI any different?

xAI has hired (and is still hiring) industry experts that have previously worked at Microsoft, OpenAI, DeepMind, and Google Research. The small team at xAI has contributed to well-known AI products like OpenAI's GPT-4 model, Deepmind's AlphaCode deep-learning model that writes software at a competitive level, and Google's Minerva language model that can solve scientific and mathematical queries with a step-by-step approach.

It's rather interesting that Musk is hiring some of the most accomplished AI experts out there, and setting up an AI lab from his own pocket, when he is also the one warning about AI's potential for civilizational damage. Musk signed a well-publicized open letter that asked for a six-month pause on the development of advanced AI models like GPT-4, citing tangible risks to humanity. Musk has also publicly criticized OpenAI on multiple occasions for reckless AI development.

Reckless is also how neuroscience experts describe Musk's effort to put a brain interface chip on a human, despite his company Neuralink falling under the scanner for animal cruelty, pathogen contamination risks, and overzealous ambitions. It would be interesting to see how xAI does things differently when calls for regulation are echoing everywhere.