Building AI to accelerate science and improve lives

构建人工智能以加速科学发展并改善人类生活

We’re asking what’s possible for health, natural disaster and weather resilience, learning, and economic opportunity. 我们正在探索人工智能在医疗健康、自然灾害与气候韧性、教育学习以及经济机遇方面所能带来的无限可能。

Google is using AI to speed up scientific breakthroughs that help people all over the world. These tools are already detecting diseases earlier, predicting natural disasters like floods and wildfires, and breaking down language barriers. They’re also helping students learn better and supporting workers as jobs change. By working with experts everywhere, Google hopes to use this technology to solve some of the world’s toughest problems. 谷歌正在利用人工智能加速科学突破，以造福全球民众。这些工具已经能够更早地检测疾病、预测洪水和野火等自然灾害，并打破语言障碍。它们还在帮助学生更有效地学习，并在就业市场变革之际为劳动者提供支持。通过与全球各地的专家合作，谷歌希望利用这项技术解决世界上一些最棘手的难题。

Today, we reached a significant milestone that stands as a testament to decades of AI research and advancement: Google technologies now support more than 300 languages, spoken by 7 billion people — representing 86% of the global population. To help us understand how these tools are driving real-world opportunity, we also released new interactive insights today with our AI & Economy ATLAS, the most comprehensive look at how real people are using AI globally. 今天，我们达成了一个重要的里程碑，这见证了数十年来人工智能的研究与进步：谷歌的技术目前已支持超过 300 种语言，覆盖 70 亿人口，占全球总人口的 86%。为了帮助我们理解这些工具如何创造现实世界的机遇，我们今天还发布了全新的交互式洞察报告——“AI 与经济地图集”（AI & Economy ATLAS），这是迄今为止关于全球民众如何使用人工智能最全面的研究。

This comes on top of a raft of key AI advances in science to benefit people over just the past few weeks: 在此基础上，过去几周我们在科学领域取得了一系列旨在造福人类的关键人工智能进展：

• We mapped all 9 billion possible single letter genetic changes across the human genome with AlphaGenome Atlas and made it openly available to researchers. 我们利用 AlphaGenome Atlas 绘制了人类基因组中所有 90 亿种可能的单字母基因变异图谱，并将其向研究人员全面开放。
• We introduced WeatherNext 3, our most advanced and accurate global weather model, delivering 50% more accurate precipitation forecasts a day or more ahead — and it’s already in use in our products. 我们推出了 WeatherNext 3，这是我们迄今为止最先进、最准确的全球天气模型，能够将提前一天或更长时间的降水预报准确率提高 50%，且该模型已应用于我们的产品中。
• We brought together data on global health, food security, and socioeconomics into a single Planetary Prediction Engine to forecast planetary crises — this has already been used in the ongoing Ebola outbreak in the Democratic Republic of the Congo and in the U.S. in identifying vulnerable communities across 21 CDC health indicators. 我们将全球健康、粮食安全和社会经济数据整合到一个统一的“行星预测引擎”（Planetary Prediction Engine）中，用于预测全球性危机。该引擎已应用于刚果民主共和国正在进行的埃博拉疫情监测，以及美国境内基于 21 项 CDC 健康指标对脆弱社区的识别工作。
• We scaled AI research to help cut the climate impact of aviation — this is already being applied in the U.K. (in collaboration with the government) and in Asia. 我们扩展了人工智能研究，以帮助减少航空业对气候的影响。该技术已在英国（与政府合作）及亚洲地区投入应用。

What ties all of this work together? It is the belief that advances in AI can accelerate scientific progress in ways that will directly improve people’s lives today and in the future. This is a key element of what motivates our work in AI. We’re focusing our work in key areas that matter most: making disease detectable, treatable, and preventable, predicting natural disasters, expanding learning, and unlocking economic opportunities for more people. 将所有这些工作串联在一起的核心理念是：我们坚信人工智能的进步能够加速科学发展，从而直接改善人们现在及未来的生活。这也是驱动我们人工智能工作的核心要素。我们正专注于最关键的领域：让疾病可检测、可治疗、可预防；预测自然灾害；普及教育；并为更多人解锁经济机遇。

While the possibilities are exciting, the benefits of AI are not guaranteed. Making them real — and mitigating their challenges and risks — demands that society works together. Though there is more still to do, AI’s progress is already making it possible for us to aspire to do bold and ambitious things that can benefit people, and to ask and address questions that were once considered impossible to solve. 尽管前景令人振奋，但人工智能带来的益处并非理所当然。要将这些愿景变为现实，并缓解其带来的挑战与风险，需要全社会的共同努力。尽管前路漫漫，但人工智能的进步已使我们能够追求那些造福人类的大胆目标，并去探索和解决曾经被认为无法攻克的难题。

We’re making progress in using AI to improve disease detection and diagnosis, and to better understand health conditions: 我们在利用人工智能改善疾病检测与诊断，以及更深入地了解健康状况方面取得了进展：

• Deepening scientific discovery: Our Nobel-Prize winning AlphaFold has predicted all 200 million protein structures known to science, providing a new basis for understanding and researching diseases. It is now used by 4 million researchers in 190 countries in areas from drug discovery to understanding neglected diseases like Chagas disease and leishmaniasis. AlphaMissense is helping researchers predict disease-causing genetic mutations. And now, we’re building on AlphaFold and AlphaMissense with AlphaGenome Atlas, offering scientists predictive insights into how genetic variations alter cellular behavior. 深化科学发现：我们获得诺贝尔奖的 AlphaFold 已经预测了科学界已知的全部 2 亿种蛋白质结构，为理解和研究疾病提供了新的基础。目前，全球 190 个国家的 400 万研究人员正在使用它，涵盖从药物研发到理解查加斯病（Chagas disease）和利什曼病（leishmaniasis）等被忽视疾病的各个领域。AlphaMissense 正在帮助研究人员预测致病基因突变。现在，我们基于 AlphaFold 和 AlphaMissense 推出了 AlphaGenome Atlas，为科学家提供关于基因变异如何改变细胞行为的预测性洞察。
• Earlier detection: Our recent breast cancer study with Imperial College London and the U.K.’s NHS showed AI can detect 25% of interval cancers previously missed in mammograms of 175,000 women. At the same time, we’re making meaningful progress in tools to help detect lung cancer, colorectal cancer, and genetic mutations in tumor cells. 更早期的检测：我们近期与伦敦帝国理工学院及英国国家医疗服务体系（NHS）合作开展的乳腺癌研究显示，人工智能能够检测出 175,000 名女性乳房 X 光检查中此前被漏诊的 25% 的间期癌。同时，我们在辅助检测肺癌、结直肠癌以及肿瘤细胞基因突变的工具方面也取得了显著进展。
• Global screenings: For tuberculosis — where ~40% of infected people worldwide go undiagnosed — our chest X-ray (used by Nexus Intelligence) has screened over 25,000 x-rays across 40 locations in six nations. We are also using bioacoustic models to detect TB via coughs using Health Acoustic Representations. Meanwhile, our diabetic retinopathy model, developed with partners, has supported more than 1.15 million screenings globally, with plans to expand to 6 million over the next decade to help detect a treatable but growing cause of preventable blindness. 全球筛查：针对结核病（全球约 40% 的感染者未得到诊断），我们的胸部 X 光检查工具（由 Nexus Intelligence 使用）已在六个国家的 40 个地点筛查了超过 25,000 张 X 光片。我们还利用生物声学模型，通过“健康声学表征”（Health Acoustic Representations）分析咳嗽声来检测结核病。同时，我们与合作伙伴共同开发的糖尿病视网膜病变模型已在全球支持了超过 115 万次筛查，并计划在未来十年内扩展至 600 万次，以帮助检测这一可治疗但日益严重的致盲原因。
• Expanding access: We’re pioneering the use of everyday smart phones and wearables for early detection of cardiovascular disease, insulin resistance, hypertension, loss of pulse, and passive heart rate monitoring. We’re working with leaders in Arkansas to help develop a blueprint for improving health outcomes in rural areas. 扩大医疗可及性：我们正在开创利用日常智能手机和可穿戴设备进行心血管疾病、胰岛素抵抗、高血压、脉搏缺失及被动心率监测的早期检测。我们正与阿肯色州的领导者合作，共同制定改善农村地区健康状况的蓝图。
• Tools for scientists and health practitioners: Collaborative AI tools like Co-Scientist are helping researchers accelerate and expand core steps of the scientific method, like generating and validating novel hypotheses (such as identifying new therapeutic applications for existing drugs for acute myeloid leukemia). We have also open-sourced AI tools like DeepConsensus, DeepVariant, and DeepPolisher. Over the last decade, these tools have assisted in completing the human genome, drafting the first pangenome, and enabling ongoing work as part of the Human Pangenome Reference Consortium, better representing human genetic diversity and allowing experts to more accurately diagnose and treat diseases. Through AMIE (Articulate Medical Intelligence Explorer), we are continuing to work on prospective evidence in real-world settings, collaborating with academic and medical institutions. 科学家与医疗从业者的工具：像 Co-Scientist 这样的协作式人工智能工具正在帮助研究人员加速并扩展科学方法的核心步骤，例如生成和验证新颖的假设（如为急性髓系白血病寻找现有药物的新治疗用途）。我们还开源了 DeepConsensus、DeepVariant 和 DeepPolisher 等人工智能工具。在过去十年中，这些工具协助完成了人类基因组测序，起草了首个泛基因组，并支持了“人类泛基因组参考联盟”（Human Pangenome Reference Consortium）的持续工作，从而更好地呈现人类遗传多样性，使专家能够更准确地诊断和治疗疾病。通过 AMIE（Articulate Medical Intelligence Explorer），我们正继续在现实环境中积累前瞻性证据，并与学术和医疗机构开展合作。