Introduction:
A custom version of ChatGPT is being developed by Elon Musk’s AI firm, xAI. Based on Musk’s late-night tweets on X on Friday, it seems that the AI model xAI has been secretly creating is true. The model, named Musk Says X Subscribers Will Get Early Access Grok (a term recently trademarked by xAI), responds to queries conversationally, potentially utilizing a knowledge base employed to train ChatGPT and other analogous text-generating models (such as Meta’s Llama 2).
According to Musk, Grok uses “real-time access” to information about X. Additionally, the model can browse the internet like ChatGPT can, which allows it to look for the most recent information on particular subjects online.
Musk hinted that Grok will decline to respond to some more delicate questions, such as “Tell me how to make cocaine, step by step.” Based on a snapshot, the model responds to that specific query in a more wry manner than ChatGPT. It’s unclear, though, whether this is a pre-programmed response or whether Musk claims in a tweet that the system is “designed to have a little more humor in its responses.”
Musk announced early on Friday that xAI would make its first AI model, likely Grok, available to a “select group” on Saturday, November 4. However, Musk stated in a later tweet that night that Grok will be available to all users of X’s newly released Premium Plus plan, which costs $16 a month for ad-free access to X, “once it’s out of early beta.”
The self-described close friend of Musk, Larry Ellison, the co-founder of Oracle, said in September that xAI had secured a contract to train its AI models on Oracle’s cloud. However, something regarding the inner workings of those AI models or the kinds of jobs they are capable of has yet to be made public by xAI.
Musk Says X Subscribers Will Get Early Access:
Musk Says X Subscribers Will Get Early Access [Source of Image: Techcrunch.com]
Musk set the lofty objective of using AI to “understand the true nature of the universe” when he announced the formation of xAI in July. Dan Hendricks, the director of the Center for AI Safety, a nonprofit organization dedicated to AI research, provides advice to the company, which is led by Musk and includes veterans from DeepMind, OpenAI, Google Research, Microsoft Research, Tesla, and the University of Toronto. The company also works with X and other businesses that support Musk, such as Tesla.
Musk stated in an April interview with Tucker Carlson that he intended to create a “maximum-truth-seeking AI.” Does Grok have AI? Maybe — perhaps it’s a first step towards something far more significant.
In a tweet on Friday afternoon, Musk was cited as saying, “It (xAI’s new model) is the best that currently exists in some important respects.”
Since the billionaire parted ways with Sam Altman and Ilya Sutskever, the co-founders of OpenAI, the company that developed ChatGPT, a few years ago, Musk’s aspirations in AI have expanded. Musk was disenchanted and even hostile to the firm. He served on the board of OpenAI as it began to prioritize commercial initiatives over open-source research. More recently, Musk severed the company’s access to X data, claiming that OpenAI was underpaying for the privilege. Musk had resigned from the OpenAI board in 2018.
