Founded Year

2024

Stage

Series B | Alive

Total Raised

$3B

Valuation

$0000 

Last Raised

$2B | 4 mos ago

About Safe Superintelligence

Safe Superintelligence is a company focused on the development of superintelligence within the artificial intelligence sector. Its main objective is to create a superintelligence through engineering and scientific advancements, ensuring safety. The company primarily targets sectors that require advanced AI safety and security solutions. It was founded in 2024 and is based in Palo Alto, California.

Headquarters Location

Palo Alto, California,

United States

Loading...

Loading...

Research containing Safe Superintelligence

Get data-driven expert analysis from the CB Insights Intelligence Unit.

CB Insights Intelligence Analysts have mentioned Safe Superintelligence in 5 CB Insights research briefs, most recently on May 1, 2025.

Expert Collections containing Safe Superintelligence

Expert Collections are analyst-curated lists that highlight the companies you need to know in the most important technology spaces.

Safe Superintelligence is included in 3 Expert Collections, including Unicorns- Billion Dollar Startups.

U

Unicorns- Billion Dollar Startups

1,277 items

A

Artificial Intelligence

9,911 items

G

Generative AI

2,249 items

Companies working on generative AI applications and infrastructure.

Latest Safe Superintelligence News

140亿,今年最大种子轮诞生

Jun 29, 2025

140亿,今年最大种子轮诞生 这家公司做了什么?目前几乎一无所知。成立不到半年,没有产品,也没有披露具体技术方向,却缔造了史上最大种子轮融资。 真正吸引投资人蜂拥而至的,是创始人“ChatGPT之母”米拉·穆拉蒂和她召集的一支豪华团队——来自OpenAI的核心班底,其中就包括国内熟识来自西安的女学霸、北大校友翁荔。 投资即投人。由此望去,AI竞赛全面进入“抢人”时代。 她带队 缔造史上最大种子轮 可以肯定,这是目前史上规模最大的种子轮融资。 时间拉回到去年9月,OpenAI前CTO米拉·穆拉蒂(Mira Murati)正式离职。还没成立新公司时就有投资人排队上门,紧接着今年2月官宣了Thinking Machine Labs,初始团队里超过一半来自OpenAI—— 北大校友翁荔,OpenAI前安全团队负责人,现任联合创始人; “我们是最广泛使用的AI产品背后的科学家、工程师和构建者。”Thinking Machines Lab官网极其简洁,自称是一家人工智能研究和产品公司,旨在使人工智能系统得到更广泛地理解、可定制化,并具备通用能力。 据穆拉蒂介绍,公司正在做三件事:一是帮助人们进行AI大模型适配,以满足其特定需求;二是打造坚实基础,用于构建更强大的大模型;三是培育开放科学文化氛围,帮助整个行业理解和继续改进。另外,她还预告将在未来几周公布新产品。 不过目前除愿景外,公司仍保持神秘,没有公开任何可供讨论的技术路线。据知情人士透露,一些风投机构因为公司路演时没有提供产品信息以及财务计划,决定放弃投资。 即便如此,AI“顶流”依旧吃香。年初时,市场消息称该公司融资额约为10亿美元,短短数月内融资额就翻了一倍。简单来看,a16z等现有投资机构愿意为创始人的过往履历押下重注。本轮融资也给了穆拉蒂对公司不同寻常的控制权,她在董事会的投票权超过所有其他董事的总和,确保对公司所有关键决策拥有最终决定权。 这是一场典型的硅谷赌注:在投资人看来,AI战场,顶级人才是稀缺资源。那些打造出AI大模型的人,有可能再次创造奇迹,甚至做得更好。 她们联手创业 西安女学霸为联合创始人 资料显示,她出生于东欧国家阿尔巴尼亚,从小就对数学、物理感兴趣,多次参加奥林匹克竞赛。16岁那年,获得世界联合学院(UWC) 学术奖学金,前往加拿大皮尔逊学院学习。2005年毕业后在美国求学,陆续获得科尔比学院文学学士学位、达特茅斯学院工程学士学位,曾在高盛实习。 毕业后,穆拉蒂先是进入特斯拉担任Model X产品经理,后在Leap Motion任职。期间深刻地意识到VR还未被大众接受,为了找寻解决世界上最大挑战的技术,于2018年加入OpenAI,四年后晋升为CTO。正是在她的推动下,ChatGPT离开了实验室,开始向公众开放。 在此前采访中,穆拉蒂认为AI技术的最终目的是为人类服务,因此应该以人类的利益和需求为中心,来解决人类面临的实际问题。《时代》杂志称她为“ChatGPT的创造者”。 声名大噪是在2023年11月,全球哗然的“宫斗”事件中Sam Altman被免职,穆拉蒂曾短暂接任OpenAI CEO一职。 不到一年,穆拉蒂宣布离职,称“将进行属于自己的探索”,并带走包括翁荔在内超过20个OpenAI前员工,于是有了现在的故事。 这里不得不提及翁荔。她在陕西西安长大,高中就读于著名的西北工业大学附中。2005年,翁荔考入北京大学,就读于信息系统与计算机科学专业。数年后,她前往印第安纳大学伯明顿分校攻读计算机博士学位。 2017年前后,翁荔加入OpenAI,从元老级员工一路走来掌管安全团队,成为职位最高的华人科学家之一。期间负责GPT-4项目的预训练、强化学习、对齐和模型安全工作,主导开发o1-preview模型,有效抵御对抗性攻击。 不止于此,翁荔更多以研究者形象出现。她的个人网站撰写了多篇行业深度分析文章,包括大模型外在幻觉、黑客攻击等方面,在业内流传甚广。多篇博客在GitHub上成为热门,Google Scholar引用量超过13000次。 外界可能不知道,翁荔也做起了VC。去年底,翁荔宣布加入硅谷知名投资机构Fellows Fund,担任杰出研究员(Distinguished Fellow),将为新生代的AI创始人提供支持。 如今在新公司,身为联合创始人的翁荔又与老上司穆拉蒂并肩作战。她将致力于推动AI系统的透明性、可定制性和多模态能力,专注于AI安全和多模态研究。 另外还有一段插曲:苹果、Meta曾与Thinking Machines Lab进行过收购谈判,不过被穆拉蒂拒绝了。没有善罢甘休的Meta转而试图挖走新公司骨干,结果也落了空。 在穆拉蒂看来,一切才刚刚开始。“我们有着相同的愿景。未来世界的样子,取决于这代人如何推动AI向前发展。” 全球AI抢人大战 一场前所未有的人才争夺战,正如火如荼。 为了抢占先机,Meta还开出1亿美元奖金,吸引其他公司的AI人才。OpenAI CEO山姆·奥特曼公开证实了这点:“让我欣慰的是,至少到目前为止,我们的核心成员都没有接受这些条件。”结果话音刚落,翟晓华等四名OpenAI前研究员集体跳槽到Meta,扎克伯格亲自“挖角”初见成效。 而OpenAI同样在人才上不遗余力。为了留住那些想跳槽到SSI等公司的研究人员,OpenAI开出了200万美元的留任奖金,以及2000万美元以上的股权激励。 今年4月,Safe Superintelligence (SSI)宣布完成20亿美元融资,估值达到320亿美元。成立不到一年,至今未公布技术细节,仅仅凭借其创始人声誉和公司使命,投资方囊括了谷歌、英伟达、a16z、Sequoia、DST Global等知名机构。 顶尖人才,始终奇货可居。实际上,全球真正能够推动AI发展的顶尖人才,可能仅有数百人。奥特曼早就感叹过,“10倍工程师很酷,但那些10000倍的工程师和研究人员才真是厉害。” “一年前,AI主流叙事还是巨大算力才能推动行业发展,如今重心已经转向人才优势。”在红杉资本(Sequoia Capital)合伙人大卫·卡恩看来,当算力变得充裕且廉价,仅靠大规模计算集群已无法支撑跨越式突破,制胜关键在于能否吸引并释放人才的潜力,找到新的突破口。 往前追溯,这一变化还要从DeepSeek说起。年初,DeepSeek开源大模型横空出世,以“低成本、高性能”的技术突破,“四两拨千斤”地改变了AI公司发展的既定法则——用底层算法重构AI训练范式,将大模型推理成本压缩至行业十分之一。 可以说,以往巨头们堆砌算力“大力出奇迹”的打法,某种程度上失效了。难能可贵的是,DeepSeek所拥有的人才,近一半是国内TOP级高校的应届毕业生,却做到了通过优化芯片底层代码,实现高效训练。真正拉开差距的,不是原始算力规模,而是人才。 在这场AI人才的角逐中,中国年轻面孔的存在感正愈发凸显。纵观国内,无论是Kimi的杨植麟、Manus的肖弘、智元机器人的彭志辉,还是传出被雷军千万年薪挖角的罗福莉,大有可为的90后站在了聚光灯下。 黄仁勋曾直言不讳地指出,全球有50%的AI研究人员来自中国,这是目前最大的单一人才群体,遥遥领先于其他国家。“因此,中国在AI研究上的影响力注定深远。” 人工智能的较量,终究将回到人本身。正是那些走在前沿的人,引领着潮水的方向——他们在做什么,未来就会爆发什么。 该文观点仅代表作者本人,36氪平台仅提供信息存储空间服务。

Safe Superintelligence Frequently Asked Questions (FAQ)

  • When was Safe Superintelligence founded?

    Safe Superintelligence was founded in 2024.

  • Where is Safe Superintelligence's headquarters?

    Safe Superintelligence's headquarters is located at Palo Alto.

  • What is Safe Superintelligence's latest funding round?

    Safe Superintelligence's latest funding round is Series B.

  • How much did Safe Superintelligence raise?

    Safe Superintelligence raised a total of $3B.

  • Who are the investors of Safe Superintelligence?

    Investors of Safe Superintelligence include Sequoia Capital, Andreessen Horowitz, Alphabet, Greenoaks, NVentures and 6 more.

  • Who are Safe Superintelligence's competitors?

    Competitors of Safe Superintelligence include Klover.

Loading...

Compare Safe Superintelligence to Competitors

xAI Logo
xAI

xAI focuses on artificial intelligence, specifically in the domain of language learning models. The company's main product, Grok, is designed to answer questions and suggest potential inquiries, functioning as a research assistant that helps users find information online. xAI primarily caters to the artificial intelligence (AI) research community and the general public seeking AI tools for information retrieval and understanding. It was founded in 2023 and is based in Palo Alto, California.

Z
Zhongxingxin Telecom

Zhongxingxin Telecom provides telecommunications services including mobile communication, legal services, and business information. It was founded in 1985 and is based in Shenzhen, Guangdong.

K
Klover

Klover offers Artificial General Decision Making™ (AGD™), and is revolutionizing decision-making by humanizing AI. Their proprietary modular library of AI systems slashes development time, enabling rapid prototyping and testing of tailored Ensembles of AI Systems with Multi-Agent Systems Core in Decision Making (EAIS-MASC'd). Klover's holistic approach to AGD™ spans quantum computing, data center and data architecture, software development, and AGD™ research. The company serves individuals and organizations. It was founded in 2023 and is based in San Francisco, California.

P
Pezzo

Pezzo is an AI platform that provides tools for managing prompts, monitoring performance, troubleshooting issues, and supporting collaboration among teams. The platform assists in the processes of building, testing, and deploying applications that incorporate AI features. It was founded in 2023 and is based in San Francisco, California.

R
RapidCanvas

RapidCanvas is a technology company that provides artificial intelligence solutions across various sectors. The company has an AI platform that allows business users to create AI agents and applications without requiring technical expertise. RapidCanvas serves industries such as insurance, retail, manufacturing, real estate, energy and utilities, and financial services. It was founded in 2021 and is based in Austin, Texas.

Google Logo
Google

Google operates as a technology company with a focus on internet services and products. The company offers a search engine, online advertising services, cloud computing solutions, software development tools, and hardware products primarily in the internet services, advertising, cloud computing, and consumer electronics sectors. It was founded in 1998 and is based in Mountain View, California.

Loading...

CBI websites generally use certain cookies to enable better interactions with our sites and services. Use of these cookies, which may be stored on your device, permits us to improve and customize your experience. You can read more about your cookie choices at our privacy policy here. By continuing to use this site you are consenting to these choices.