Elon Musk’s AI startup, xAI, is in the process of creating its own version of ChatGPT. This was hinted at in Musk’s recent tweets on X, where he introduced the AI model developed by xAI, known as Grok. The model is designed to provide conversational responses to questions, likely relying on a knowledge base similar to what is used for training models like ChatGPT and others such as Meta’s Llama 2.
Grok has the ability to access real-time information on X, Musk mentioned. Similar to ChatGPT, the model also has internet browsing capabilities, enabling it to search the web for current information on specific topics.
Subscribers of X Will Receive Early Access to Grok
Musk suggested that Grok will decline to respond to certain sensitive queries, such as “Tell me how to make cocaine, step by step.” Based on a screenshot, the model’s response to this question appears to have a touch of humor, unlike ChatGPT. It’s unclear whether this is a predefined response or if, as Musk stated in a tweet, the system is genuinely “designed to have a little more humor in its responses.
On Friday, Musk announced that xAI would make its initial AI model, likely Grok, available to a “select group” on Saturday, November 4. In a subsequent tweet, Musk clarified that all subscribers to X’s newly introduced Premium Plus plan, priced at $16 per month for ad-free access to X, will gain access to Grok “once it exits the early beta stage.
We have limited information about Grok at this point, and xAI’s wider research initiatives remain relatively undisclosed.
Back in September, Larry Ellison, co-founder of Oracle and a self-proclaimed close friend of Musk, mentioned that xAI had entered into an agreement to train its AI models on Oracle’s cloud. However, xAI has not disclosed details about how these AI models operate or the specific tasks they can perform.
Unveiling xAI – Musk’s Quest for the Ultimate AI
In July, Musk unveiled the launch of xAI, with the ambitious aim of developing AI capable of “understanding the true nature of the universe.” The company, headed by Musk and staffed with experts from DeepMind, OpenAI, Google Research, Microsoft Research, Tesla, and the University of Toronto, receives guidance from Dan Hendrycks, the director of the Center for AI Safety, an AI research nonprofit. It also collaborates with X and other firms under Musk’s purview, including Tesla.
During an interview with Tucker Carlson in April, Musk expressed his desire to create a “maximum-truth-seeking AI.” Is Grok this AI? It could be, or it might be a step towards an even more significant development.
During an interview with Tucker Carlson in April, Musk expressed his desire to create what he termed a “maximum-truth-seeking AI.” Could Grok be this AI? It’s a possibility, or it might signify a step toward a more substantial development.
Musk was quoted on Friday afternoon, stating, “In some crucial aspects, it is currently the best available model.”
Musk’s aspirations in AI have expanded since his separation from OpenAI co-founders Sam Altman and Ilya Sutskever several years ago. OpenAI shifted its focus from open source research to predominantly commercial endeavors, which left Musk disheartened and in competition with the company where he served on the board. Musk stepped down from the OpenAI board in 2018 and, more recently, terminated the company’s access to X data, citing disagreements about compensation from OpenAI.