加速人工智能在新加坡的影响
内容总结：
新加坡深化AI国家战略布局 谷歌签署新一轮合作聚焦健康、教育与安全
作为全球数字化转型的典范，新加坡正通过其国家人工智能战略，系统性推进人工智能的大规模部署。近日，谷歌与新加坡政府宣布达成新一轮国家级AI合作伙伴关系。该合作由新加坡数码发展及新闻部牵头，联合多个公共机构与部门，旨在将前沿AI技术转化为社会福祉的强大动力。
在合作框架下，双方将重点攻克关键社会挑战，培育适应AI时代的劳动力，并推动企业创新以构建AI驱动型经济，同时确保信任与安全置于核心位置。
加速生命科学与健康领域突破
依托谷歌DeepMind在新加坡设立的实验室，双方将联合推动高影响力领域的研发。在医疗健康方面，DeepMind正探索与当地公共卫生集团合作开展“AI协同临床研究”项目，利用AI辅助医生提升诊疗质量。这包括开发能在医生临床授权下为患者提供全程支持的AI代理，并从临床指南和科学文献中精准提取信息。
为加速科学发现，谷歌将协助新加坡国家研究基金会培训本地科研人员使用其AI for Science工具（如“Co-Scientist”），该工具已在生物医学领域展现潜力。同时，谷歌还将为新加坡科技研究局（A*STAR）的研究人员提供基于谷歌云的安全AI工具，支持其在受控环境中分析科学数据并加速新发现。
在包容性创新方面，谷歌DeepMind正为视障运动员开发一款跑步代理工具。该工具利用实时空间推理技术，帮助运动员在无需物理引导线或人工引导的情况下独立跑步。谷歌已与新加坡残疾与包容事务主管机构合作，对产品进行实际测试与迭代。
赋能教育与未来劳动力培养
为确保AI惠及全民，谷歌正与政府紧密合作构建面向未来的教育生态系统。目前，谷歌已为新加坡从小学到初级学院的所有教育工作者启用了Google Workspace for Education中的高级AI功能，帮助教师安全地使用AI辅助备课和定制教材。
在此基础上，新加坡教育部与谷歌将扩大合作，强化教育部在教与学方面的AI能力，包括教师培训和技能提升计划。这些举措是新加坡整体评估企业解决方案在支持教育成果方面应用与规模化的一部分。
驱动企业创新增长
为促进企业和初创生态系统的创新增长，谷歌正在扩大其谷歌云新加坡工程中心的前沿部署工程师团队，以帮助本地企业加速并规模化代理式企业转型。
筑牢AI安全与信任基石
为实现AI的全面收益，安全必须置于首位。该合作旨在将新加坡打造为信任技术的全球标杆，发展负责任AI部署所需的安全框架和工具。
针对日益强大的代理系统（特别是“计算机使用”代理），谷歌与新加坡网络安全局、政府科技局及资讯通信媒体发展局联合发布白皮书，分享AI代理沙盒测试的发现与建议，包括确保代理能安全高效执行软件测试、社会救助申请等任务的最佳实践。
此外，谷歌DeepMind正与新加坡资讯通信媒体发展局及MLCommons合作，研究多模态和多语言安全基准。该合作旨在支持尊重本地语言和文化细微差别的AI安全部署，确保数字未来真正为大众共建共享。
展望未来
实现这一宏伟愿景需要深度且可信赖的合作。谷歌表示，将继续与新加坡政府、企业和社区携手，共同构建一个安全、创新且惠及全民的未来。
中文翻译：
加速人工智能在新加坡的影响
新加坡长期以来一直是全球数字创新的典范。通过其国家人工智能战略，该国正在积极规划如何大规模部署人工智能——不仅是为了提升经济竞争力，更是为了切实改善日常生活。
今天，我们通过与新加坡政府建立新的国家人工智能合作伙伴关系，深化了在新加坡的人工智能投资。这项由数字发展与信息部领导的新合作，汇集了多个公共机构和部委，共同将前沿人工智能作为推动社会进步的重要力量。
我们携手扩大承诺，致力于解决关键社会挑战、培养适应人工智能时代的劳动力、推动企业创新以打造人工智能驱动的经济，同时始终将信任与安全置于首位。
加速健康与生命科学领域的突破
人工智能正在加速解决人类面临的重大挑战，而新加坡是创新人工智能解决方案的理想之地。借助谷歌深度思维研究实验室在新加坡的布局及其更广泛的国家人工智能合作伙伴计划，我们正与当地专家合作，在以下高影响力领域推动研发：
在健康领域，谷歌深度思维正在探索与公共卫生集群合作，开展其全球“人工智能协同临床研究”计划，研究如何通过人工智能增强和支持医生的专业知识，以提供更高质量的护理。这包括在医生临床授权下，通过人工智能代理为患者提供全流程护理支持，这些系统能够从临床指南和科学文献中提取更精确的信息。
为加速科学发现，我们还与新加坡国家研究基金会合作，培训当地科学界使用我们的智能科学工具（如“协作科学家”），这些工具已在生物医学领域展现出潜力。此外，还将举办研讨会，帮助科学界掌握这些前沿工具的使用方法。
这一合作同样延伸到新加坡科技研究局。我们将为其研究人员和员工提供谷歌云安全且具备人工智能功能的工具（包括假设生成能力），以支持科学研究和分析。这使他们能够在受控环境中安全分析科学数据，加速新发现的同时保护知识产权。
真正的创新需要包容性。为此，谷歌深度思维正在为视障运动员开发一款跑步辅助代理。该工具利用实时空间推理技术，帮助运动员无需物理引导线或人类向导即可独立跑步。我们正与新加坡负责残疾与包容事务的机构合作，测试并迭代该产品，以满足视障跑者的实际需求。
助力教育工作者并培养未来劳动力
为确保人工智能惠及所有人，我们正与政府紧密合作，构建面向未来的教育生态系统。
基础已经奠定。谷歌已为从小学到初级学院的所有教育工作者启用了谷歌工作空间教育版的高级人工智能功能。这为教师提供了安全的人工智能辅助工具，如课程规划与教材定制，使他们有更多时间专注于教学和学生指导。
在此基础上，新加坡教育部与谷歌正扩大合作，以增强教育部在教与学中的人工智能能力，包括教师培训与技能提升项目。这些努力是教育部更广泛战略的一部分，旨在评估企业解决方案如何有效应用并规模化推广，以支持教育成果。
这些举措建立在谷歌新加坡持续开展的“前进吧！人工智能”倡议之上。该倡议整合了高接触性与规模化项目，旨在赋能新加坡各阶层民众。
以创新驱动增长
为推动企业和本地创业生态系统的创新增长，我们正扩大谷歌云新加坡工程中心的前沿部署工程师团队，帮助新加坡企业加速并规模化推进智能代理驱动的企业转型。
确保新加坡人工智能生态系统的安全与可信
最后，要充分发挥人工智能的潜力，安全必须放在首位。此次合作旨在将新加坡打造为全球信任技术领域的标杆，制定负责任人工智能部署所需的安全框架和工具。
随着智能代理系统（特别是“计算机使用”代理）能力不断增强，我们需要理解它们在现实世界中的行为。谷歌、新加坡网络安全局、新加坡政府科技局及资讯通信媒体发展局联合发布的白皮书，探讨了人工智能代理沙盒的发现与建议，包括这些代理安全高效执行软件测试和社会援助申请等任务所需的最佳实践。
此外，谷歌深度思维正与资讯通信媒体发展局及MLCommons合作，研究多模态和多语言安全基准。这一合作专注于支持尊重当地语言与文化细微差别的安全、负责任人工智能部署，帮助我们构建一个为所有人设计、与所有人共创的数字未来。
展望未来
实现这一宏伟愿景需要深度且可信赖的合作。我们无比自豪能与新加坡政府、企业和社区继续同行，共同构建一个安全、创新且惠及所有人的未来。
英文来源：
Accelerating AI impact in Singapore
Singapore has long been a global blueprint for digital innovation. Through its National AI Strategy, the nation is actively charting a course for how AI can be deployed at scale — not just to fuel economic competitiveness, but to meaningfully improve everyday lives.
Today, we’re deepening our AI investments in Singapore through a new national AI partnership with the Singapore government. Led by the Ministry of Digital Development and Information (MDDI), this new partnership brings together multiple public agencies and ministries to harness frontier AI as a powerful force for good.
Together, we are expanding our commitment to solve critical societal challenges, foster an AI-ready workforce, and drive enterprise innovation for an AI-driven economy, while keeping trust and safety at the forefront.
Accelerating breakthroughs in health and life sciences
AI is accelerating progress on some of humanity's biggest challenges, and Singapore is the ideal place to innovate on new AI solutions. Supported by the presence of the Google DeepMind research lab in Singapore and part of its broader National Partnerships for AI initiative, we are working alongside local experts to drive research and development on high-impact areas:
In health, Google DeepMind is exploring a collaboration with public health clusters in its global “AI co-clinician research” initiative to explore how AI can augment and support a doctor’s expertise to deliver higher quality care. This includes AI agents to support patients throughout their care journeys under the clinical authority of their physician, with systems that can provide more precise information sourced from clinical guidelines and scientific literature
To accelerate scientific discovery, we are also collaborating with the National Research Foundation (NRF) to train local scientific communities to use our agentic AI for Science tools like Co-Scientist, which are already showing promise in a range of biomedical applications. This will be complemented with workshops to help the scientific community use these frontier tools.
This also extends to our partnership with Agency for Science, Technology and Research (A*STAR) where we’ll equip researchers and staff with Google Cloud’s secure, AI-enabled tools including hypothesis generation capabilities to support scientific research and analysis. This allows them to securely analyze scientific data within a governed environment, and speed up new discoveries while safeguarding its intellectual property.
True innovation requires inclusivity. To do that, Google DeepMind is developing a running agent for blind and low vision athletes. Using real-time spatial reasoning, this tool helps the athletes run independently without physical lines or human guides. We are partnering with Singapore’s focal agency for disability and inclusion—to test and iterate the product so it meets the real-world needs of vision-impaired runners.
Helping educators and training tomorrow's workforce
To ensure the benefits of AI reach everyone, we are working closely with the government to build a future-ready education ecosystem.
The foundation is already in place. Google has enabled advanced AI functionalities within Google Workspace for Education for all educators from primary schools to junior colleges. This provides teachers with secure, AI-powered assistance like planning lessons and tailoring course material, giving them more time to focus on teaching and mentoring students.
Building on this, the Ministry of Education (MOE) and Google are expanding their collaboration to strengthen MOE’s AI capabilities across teaching and learning, including educator training and upskilling programmes. These efforts are part of MOE’s broader approach to evaluate how enterprise solutions can be applied effectively and scaled in support of education outcomes.
These efforts build on Google Singapore’s ongoing Majulah AI initiative, which brings together high-touch and scaled initiatives to empower every segment of the population in Singapore.
Driving innovation for growth
In its efforts to drive innovations for growth for enterprises and the local startup ecosystem, we are expanding the team of Forward Deployed Engineers (FDEs) at the Google Cloud Singapore Engineering Center, to help Singapore-based companies further accelerate and scale agentic enterprise transformation.
Keeping Singapore’s AI ecosystem safe and trusted
Finally, to realize the full benefits of AI, safety must come first. This partnership aims to establish Singapore as a global lighthouse for trust technologies, developing the safety frameworks and tools necessary for the responsible AI deployment.
As agentic systems — specifically "computer use" agents — become more capable, we need to understand how they behave in the real world. A joint whitepaper developed by Google, the Cyber Security Agency of Singapore (CSA), Government Technology Agency of Singapore (GovTech Singapore), and the Infocomm Media Development Authority (IMDA) discusses findings and recommendations from their AI Agents Sandbox. These include the best practices necessary for these agents to safely and efficiently perform tasks like software testing and social assistance application.
Additionally, Google DeepMind is collaborating with IMDA and MLCommons to research multimodal and multilingual safety benchmarks. This collaboration is focused on supporting the safe and responsible deployment of AI that respects the nuance of local languages and cultures, helping to ensure that the digital future we build is one that is designed for and with everyone.
Looking ahead
Realizing this bold vision requires deep, trusted collaboration. We are incredibly proud to continue this journey alongside Singapore's government, businesses and communities to build a future that is safe, innovative and beneficial for all.