OpenAI’s anticipated release of its first open-weights model since GPT-2 has been delayed due to a safety review, as stated by CEO Sam Altman. This prompts concerns regarding the US’s position in developing open AI models, particularly given the mixed reception of Meta’s Llama 4 and its recent struggles. While several companies, such as Microsoft and IBM, have released smaller models, progress has largely remained behind closed doors, in stark contrast to China’s advancements. Chinese companies like DeepSeek and MiniMax have unveiled impressive open-weights models, featuring novel architectures and massive parameter counts. Despite US efforts, Chinese developers continue to excel, fostering a competitive edge in generative AI. With OpenAI’s upcoming model, originally slated for earlier this year, developers hope for renewed competition in the open model space. However, concerns persist as Meta and xAI pivot towards closed-source developments. As the year unfolds, the quest for impactful open AI continues.
Source link
