Elon Musk’s xAI company officially released its new AI models, Grok-2 and Grok-2 mini, into beta. They are only accessible with a Premium or Premium+ subscription on the X platform (formerly Twitter). Actually, this is the major upgrade from its predecessor, Grok 1.5 model, since it can carry out more reasoning and finally drop images directly on the social media page.

In a playful twist of AI irony, this image about Grok-2
was brought to life by its friendly rival, DALL-E.

Grok-2 is one of the advanced models that both generate text and visual content. This model could answer user prompts with images, unlike other models such as OpenAI’s DALL-E or Google’s Gemini.

Seemingly, Grok-2 was coming up with fewer constraints on the type of images it could generate, which broadened the range of the output, including potentially controversial or politically sensitive content. This has led to discussions on the ethics involved in such processes, specifically the model’s ability to give images that relate to well-known political personalities and copyright-related characters.

Another model, the Grok-2 mini, is a smaller version developed for faster answers with the best quality of outputs. Both are part of the larger strategy xAI is pursuing to enhance AI interactions on the X platform, making them useful across a vast array of tasks ranging from simple queries to more complex creative processes.

Source: https://x.ai/blog/grok-2

Read AI Briefings