Social network X is introducing Grok, the chatbot from xAI, to its Premium tier subscribers. This expansion comes after an announcement by Elon Musk to open up access to more paid users last month. The company has stated on its support page that only Premium and Premium+ users will be able to interact with the chatbot in select regions.
Last year, Musk’s xAI launched Grok, making it available exclusively to Premium+ users – individuals who subscribe for $16 per month or $168 per year. However, with the latest update, the chatbot can now be accessed by users paying only $8 per month.
Users can engage with Grok in either “Regular mode” or “Fun mode”. As with any other Large Language Model (LLM) product, Grok displays warnings indicating that the chatbot may provide inaccurate responses.
We’ve already witnessed some instances of this. Just this week, X introduced a new explore view within Grok, where the chatbot offers summarizations of trending news stories. Notably, both Jeff Bezos and NVIDIA-backed Perplexity AI also provide news story summarizations.
Grok now provides summarized versions of all the trending news and topics. The explore page and Grok’s home screen offer easy access to this feature. DogeDesigner (@cb_doge) April 5, 2024
However, Grok appears to go beyond just summarizing news stories, as it can also generate headlines. Mashable reported that the chatbot wrote a fake headline claiming “Iran Strikes Tel Aviv with Heavy Missiles.”
It’s likely that Musk seeks to increase the number of users utilizing Grok in order to compete with other similar products, such as OpenAI’s ChatGPT, Google’s Gemini, or Anthropic’s Claude. In the past few months, he has publicly criticized OpenAI and their operations. In fact, in March, he even filed a lawsuit against the company, claiming they had “betrayed” their non-profit goal. As a response, OpenAI released emails between Musk and their CEO, and filed papers seeking the dismissal of all claims against them.
Last month, xAI made Grok open-source, although without any documentation on the training data used. As my colleague Devin Coldewey argued, there are still many unanswered questions about the model and whether xAI will be more transparent in its development process and release information about the training data.
Your article helped me a lot, is there any more related content? Thanks!