Google's DeepMind division is reportedly paying some of its former employees an year worth's of salary to prevent them from ...

Google DeepMind predicts that artificial general intelligence could arrive in five years from now and "permanently destroy ...

近日，谷歌旗下DeepMind实验室最新发布了长达145页的人工智能安全报告，称具备人类智能水平的通用人工智能（AGI）可能在2030年左右问世，并可能造成“永久毁灭人类”的严重后果。

为了在竞争激烈的 AI 领域保持竞争优势，Google DeepMind 采取了推迟发表 AI 论文的策略。诺奖得主 Sir Demis Hassabis 领导的团队引入了更严格的审查程序和更多的官僚主义，增加了 AI ...

在竞争愈发激烈的人工智能领域，Google DeepMind选择了一条不同寻常的道路——推迟发布其AI研究论文。由诺奖得主Sir Demis Hassabis领导的团队，实施了更加严格的审查程序，造成AI论文的发表变得尤为艰难。此举显然是为了保护其在高科技行业的竞争优势，避免因过早披露技术细节而被竞争对手利用，或被暴露出其Google Gemini AI模型的潜在劣势。

FT爆料称，由Demis Hassabis领导的谷歌AI部门，正对研究论文发布实施更严格的审查机制。 2017年，「Attention is All you ...

Researchers at Google DeepMind have shared risks associated with AGI and how we can stop the technology from harming humans.

Google DeepMind staff accuse the tech giant of paying them not to join rivals for up to a year after leaving the company., ...

只需使用一种通用算法，就可以解决来自各个应用领域的各种任务，一直是人工智能（AI）行业的基本挑战之一。 如今，Google DeepMind 在这一方向取得了新的突破。 他们开发的第三代 Dreamer 通用算法，只需一次配置，就能在 150 ...

Demis Hassabis, a childhood chess prodigy, said his experience with the game afforded him his first brush with AI. He was ...

DeepMind developed a framework that highlights areas where adversarial AI is weakest, allowing defenders to prioritize ...