去到今時今日，AI 已步入不少人的日常生活中，當中除了近期大熱的 DeepSeek 外，相信由 OpenAI 開發的 ChatGPT 亦是更多用家日常會用得著的工具。OpenAI 的 CEO Sam Altman 昨日就透過 X 表示，會在月內推出自 GPT-2 誕生以來首個開放權重（Open-weight）AI 模型，
The announcement marks a strategic shift by OpenAI, which until now has been a fierce defender of closed, proprietary models that do not allow developers to modify the basic technology
In context: The constant improvements AI companies have been making to their models might lead you to think we've finally figured out how large language models (LLMs) work. But nope – LLMs continue to be one of the least understood mass-market technologies ever. But Anthropic is attempting to change
The company launched Qwen2.5-Omni-7B on Thursday as the latest addition to its Qwen family of models. With just 7 billion parameters, it is designed to run on mobile phones, tablets and laptops
繼 去年 公佈首個具推理能力的Gemini 2.0 Flash Thinking模型後， Google昨（25）日發表Gemini 2.5模型 ，作為Gemini新一代模型家族的最新成員，Google宣稱在多項測試中擊敗OpenAI o3-mini、DeepSeek R1等模型。
Gemini 2.0 Flash Thinking和Gemini 2.5是結合強化式學習及思
