Boardroom breaks down what Grok is, how it’ll be integrated into X’s platform, and how it differs from other AI chatbots on the market.
Elon Musk unveiled a new artificial intelligence system called Grok, a conversational AI chatbot that responds humorously.
Musk’s AI company, xAI, developed the new generative AI product, which is the first tech prototype the startup has built. Grok is only in its first two months of training and is currently in a beta testing phase with a small group of users.
Musk has some work to do if he plans to put Grok up against flagship platforms like OpenAI’s ChatGPT. Here’s everything you need to know about Grok and what’s planned for the new AI chatbot.
![](https://sp-ao.shortpixel.ai/client/to_auto,q_glossy,ret_img/https://boardroom.tv/wp-content/uploads/2023/08/Boardroom_TechTalkFin2.jpg)
How Grok Works
In xAI’s announcement, the company said it modeled Grok after “the Hitchhiker’s Guide to the Galaxy” and gave it a bit of Musk’s personality to answer queries with sarcastic responses that he might say himself. The AI startup aims to build a “powerful research assistant for everyone” by providing relevant information, new ideas, and data. Grok-1 is the AI chatbot’s first public-facing product, but xAI had an earlier model called Grok-0 that was developed just for training purposes.
Much like other AI chatbots on the market, users have to prompt Grok with questions to illicit AI-powered replies. Grok is trained on various datasets, including real-time information from X, even though the platform is infested with misinformation. It’s unclear what Grok’s fact-checking process is, but like other AI chatbots, xAI lets users know that Grok “can still generate false or contradictory information.”
Musk has been previewing what the AI chatbot could do. In this particular post, the tech CEO was trying to demonstrate how Grok is ahead of the game by leveraging real-time data, but his post was quickly tagged with a community note about how the varying platforms were prompted.
Here is another example Musk shared on how Grok responds.
Grok’s early launch comes as Musk announced that xAI will fold into X, bringing his AI operation under the social media network’s umbrella. xAI only launched earlier this year with a mission “to advance our collective understanding of the universe” by “building artificial intelligence to accelerate human scientific discovery,” the company’s website reads. The AI startup’s merger with X brings former employees from DeepMind, OpenAI, Google, Microsoft, and Tesla to the platform.
As for the platform’s name, the term Grok was created by science fiction author Robert A. Heinlein and first used in his novel Stranger in a Strange Land in 1961. Heinlein didn’t officially define the term in his book, but the term was used in the Martian language to describe someone who is deeply intuitive and understanding. xAI filed a trademark application for the term Grok at the beginning of November, but it’s unclear if the company was granted the trademark and if Heinlein’s estate has any say in the matter.
What’s Next?
Grok won’t be available to the masses once it hits a wider release. For now, interested users have to be verified on X to gain access to the beta version, and this requirement will likely carry over once Grok officially rolls out. Grok is still accepting users on its waiting list for its beta testing period. xAI will be taking feedback from users testing Grok-1 and implementing changes before doing a wider launch.
Once Grok rolls out, the AI chatbot will become a feature of X Premium+, X’s highest-tiered subscription service that costs $16/month.
Since Grok isn’t going to have a version available to everyone, it’ll be interesting to see if the AI chatbot can attract as many users as ChatGPT, which has an open-access free version.