Stability AI Releases Open-Source Chatbot While AI Art Wins Major Photo Award
Stability AI introduces StableLM and Stable Vicuna as free, open-source alternatives to proprietary language models. Meanwhile, a photographer deliberately wins a prestigious photo competition with an AI-generated image to expose a lack of industry regulations.
Stability AI expands its open-source offerings by releasing StableLM, a new suite of large language models trained on 3 billion and 7 billion parameters. The company follows up this initial launch with Stable Vicuna, which stands out as the first open-source chatbot trained using reinforcement learning from human feedback. These models generate both text and code, performing impressively well on conversational tasks despite their relatively small size, and developers access them freely for commercial or research use under a permissive license.
In a bold move that sparks global debate, Berlin-based artist Boris Eldagsen refuses the first prize in the Sony World Photography Awards after winning with an AI-generated image. He submits the artwork to the creative category on purpose to test whether prestigious competitions are prepared to handle artificial intelligence. By declining the award on stage in London, Eldagsen highlights the urgent need to distinguish between traditional photography and AI-generated imagery.
Eldagsen's provocative stunt exposes significant gaps in how the art world currently regulates and categorizes machine-generated content. He argues that AI images are entirely separate entities that should not compete against human-captured photographs in the same categories. This high-profile event forces the creative industry to confront difficult questions about authorship, authenticity, and the future definition of a photograph in the age of artificial intelligence.