Grok-2 and Grok-2 mini, the latest AI models developed by Elon Musk-owned xAI, have been launched in beta. These models come with improved reasoning capabilities and have the ability to generate images on the X social network. However, access to Grok is currently limited to Premium and Premium+ users on X.
According to xAI’s blog post, Grok-2 is a significant improvement over its predecessor, Grok-1.5, and boasts advanced capabilities in chat, coding, and reasoning. Alongside Grok-2, xAI has also introduced Grok-2 mini, a smaller but capable sibling. The company has already tested an early version of Grok-2 on the LMSYS leaderboard under the name “sus-column-r.”
In the coming month, xAI plans to make both Grok-2 and Grok-2 mini available to developers through its enterprise API. This move will allow developers to harness the power of these models for their own applications and projects.
While the image generation feature of Grok has been well-received by users, there are concerns regarding its potential misuse. Early images generated by users suggest that Grok doesn’t have any guardrails when it comes to creating images of political figures. With the U.S. presidential election approaching, xAI may face pressure to add limits on these features to prevent the spread of misinformation.
It is worth noting that the images generated by Grok are created using FLUX.1 by Black Forest Labs, as indicated by the text below the sample image prompts. However, it remains unclear whether these AI-generated images contain any metadata to indicate their origin.
Despite the lack of specific details about Grok 2’s capabilities, app researcher Nima Owji claims that it excels in code generation, writing, and news. However, there isn’t much evidence to support these claims. It should be mentioned that Grok’s first version has made mistakes in the past when summarizing news.
The potential for misuse of Grok’s image generation feature raises concerns about the spread of misinformation on X and other platforms. It is imperative for xAI to address this issue and implement measures to limit image generation for harmful purposes. However, it remains uncertain how X plans to tackle this challenge, as the company has been unresponsive to media inquiries since Musk’s takeover.
On a positive note, xAI has announced its plans to deploy Grok-2 and Grok-2 mini in AI-driven features on X. This includes improved search capabilities, post analytics, and reply functions. This suggests that X might introduce AI-powered replies, enhancing user interactions on the platform. Additionally, xAI will release a preview of multimodal understanding as a core part of Grok’s experience on X and API, further expanding the capabilities of the AI models.
