News on Medial

Musk’s new Grok upgrade allows X users to create largely uncensored AI images

Arstechnica · 7m

Elon Musk's AI company xAI has released the beta version of two new language models, Grok-2 and Grok-2 mini, to X subscribers. These models are integrated with the Flux image synthesis model, allowing users to create photorealistic images on the platform. However, concerns have been raised about the potential for controversial and fake images generated by Grok, as it lacks safeguards and watermarks. The responsibility for misuse of AI image generation remains a question yet to be resolved. xAI claims that Grok-2 and Grok-2 mini provide significant advancements in language model capabilities, with Grok-2 demonstrating higher performance in benchmarks. However, the GPT-4 barrier has not been surpassed yet.