The Rundown AI -每日精选 前天 15:33
Mira Murati's $12B startup sets launch clock
index_new5.html
../../../zaker_core/zaker_tpl_static/wap/tpl_guoji1.html

 

前OpenAI首席技术官Mira Murati创立的Thinking Machine Labs AI公司,在公司成立不到一年且尚未发布产品的情况下,成功获得了20亿美元的种子轮融资,公司估值飙升至120亿美元。该公司正致力于开发具有“重大开源组件”的多模态AI,并计划在未来几个月内发布首款产品。该产品将通过自然交互方式与用户协作,并可能专注于帮助企业提升利润。此举在竞争激烈的AI领域备受关注,预示着AI行业的又一重要参与者即将崛起。

💡 Thinking Machine Labs AI获20亿美元种子轮融资,估值达120亿美元,显示出市场对其技术和愿景的高度认可。该公司由前OpenAI高管Mira Murati创立,尽管尚未发布产品,已吸引大量投资。

🚀 公司致力于开发多模态AI,并将包含“重大开源组件”,旨在为研究人员和初创企业构建定制模型提供支持。产品将通过对话和视觉等自然交互方式与用户协作。

📈 Thinking Machine Labs AI的首款产品预计将在未来数月内发布,其潜在应用方向包括帮助企业提高利润。鉴于其在AI领域的背景和巨额融资,其产品发布备受期待。

🎬 此外,文章还提到了Runway公司推出的Act-Two,一款下一代AI运动捕捉模型,能够将单人表演视频转化为全动画角色,并已与多家好莱坞公司达成合作,推动AI在电影制作中的应用。

🤝 AI研究人员正呼吁加强对AI推理过程“思维链”的监控,认为这是保障AI安全的重要工具,并提议开发标准化评估方法,以应对未来模型可能出现的透明度下降问题。

Read Online | Sign Up | Advertise

Good morning, {{ first_name | AI enthusiasts }}. Former OpenAI CTO Mira Murati has kept her new startup quiet since launching in February — but that didn’t stop investors from betting a fresh $2B on whatever she's building.

With Thinking Machine Labs now valued at $12B and promising multimodal AI with a “major open-source component” in the coming months, the countdown to their first product has become one of the AI world's most anticipated reveals.

Member note: The Rundown has secured 500 private invitations for Perplexity Comet. Unique codes are available now for full members of The Rundown University.


In today’s AI rundown:

    Thinking Machine Labs raises $2B, nears product launch

    Runway’s Act-Two for AI motion capture

    Automate personalized sales outreach at scale

    AI researchers unite on reasoning transparency

    4 new AI tools & 4 job opportunities

LATEST DEVELOPMENTS

THINKING MACHINES LAB

💰 Thinking Machine Labs raises $2B, nears product launch

Image source: Getty Images

The Rundown: Former OpenAI CTO Mira Murati just announced a new $2B seed raise for her startup, Thinking Machines Lab AI, also revealing that the secretive company will share its first product in “the next couple of months.”

The details:

    The $2B seed round brings the company’s value to $12B, less than a year after its creation, with no product and little public information on direction.

    Murati said the startup’s first product will feature “a major open-source component” for researchers and startups building custom models.

    She also revealed the lab is building multimodal AI that collaborates with users in natural interactions via conversation and sight.

    The Information recently reported that TML is planning to develop custom AI models to help businesses increase profits.

Why it matters: With fierce competition among the current AI leaders, it’s easy to forget the two former OpenAI heavyweights in TML and SSI are still in stealth mode. Despite little info or product, Murati has had no shortage of investor (and acquisition) interest — making the startup’s first release a big one to watch in the coming months.

TOGETHER WITH PREZI

⭐️ Your best presentation yet, built by AI

The Rundown: Prezi AI turns boring decks into dynamic presentations that get results in meetings, pitches, and high-stakes boardrooms. Prezi has helped presenters for 15 years—and with AI, it’s even faster. No design skills needed.

With Prezi, you can:

    Go from rough ideas or PDFs to polished presentations in seconds

    Stand out with visuals that hold attention and drive real engagement

    Get AI-powered suggestions for content, structure, and design

Try Prezi AI for free — see why 160+ million already have.

RUNWAY

🎬 Runway’s Act-Two for AI motion capture

Image source: Runway

The Rundown: Runway just launched Act-Two, the company’s next-gen motion capture model that translates single performance videos into fully animated characters with head, face, body, and hand tracking across artistic styles and outputs.

The details:

    The system captures subtle facial expressions, upper body movements, hands, and backgrounds from a single driving performance video.

    Requiring just a single character reference photo, Act-Two animates and maps the driving video while maintaining backgrounds and art styles.

    Runway claims the model delivers major performance gains over October 2024’s Act-One release, particularly in consistency, fidelity, and movement.

    The company has inked partnerships with Hollywood players like Lionsgate and AMC Networks, pushing to further infuse AI into filmmaking workflows.

Why it matters: While AI adoption in Hollywood is publicly frowned upon, it’s privately being adopted en masse — and few companies are pushing that shift harder than Runway. With creative models continuing to evolve at a wild pace, AI’s use in the industry is coming, whether fans and actors want it or not.

AI TRAINING

📧 Automate personalized sales outreach at scale

The Rundown: In this tutorial, you will learn how to create a Zapier AI agent that automatically researches companies from your spreadsheet and drafts personalized sales emails, saving hours of manual outreach.

Step-by-step:

    Create a new Zapier agent and set the trigger to “On demand”

    Write detailed instructions: research company → populate contact info → draft personalized email → generate subject and body

    Connect the Google Sheets tool and select your leads spreadsheet with company, name, and email columns

    Add Gmail integration with “Create Draft” option (not send) for review control

    Test with a small batch, review generated drafts, then scale to the full lead list.

Pro tip: Always use “Create Draft” instead of “Send Email” so you can review and refine each message before it goes out.

PRESENTED BY IBM

🤖 Agentic AI for APIs

The Rundown: IBM’s API Agent brings agentic AI to API Connect, transforming prompts into production-ready APIs.

With API Agent, you can:

    Let AI handle design, validation, and deployment to Code Engine automatically

    Manage standards and cut manual steps for fast API delivery

    Harness agentic AI to turn your API ideas into secured, production-ready reality

Try API Agent today.

AI RESEARCH

🧠 AI researchers unite on reasoning transparency

Image source: Ideogram / The Rundown

The Rundown: Leading researchers from OpenAI, DeepMind, Anthropic, and other major institutions published a paper calling for deeper investigation into monitoring the "chains of thought" in AI reasoning models, arguing it could be a vital safety tool.

The details:

    The paper highlights “chain-of-thought” (CoT) traces, the model’s step-by-step problem-solving paths, as a rare window into model decision-making.

    The researchers call for a deeper study of tracking these reasoning processes, warning that transparency could erode as models evolve or training shifts.

    Notable signatories include OpenAI's Mark Chen, SSI's Ilya Sutskever, Nobel laureate Geoffrey Hinton, and DeepMind co-founder Shane Legg.

    Researchers propose developing standardized evaluations for "monitorability" and incorporating these scores into deployment decisions for frontier models.

Why it matters: This rare moment of industry consensus highlights a critical inflection point where we can still see AI's decision-making process — and the concern that competitive pressures may eventually lead to that window being closed if AI labs don’t act to preserve it.

QUICK HITS

🛠️ Trending AI Tools

    🤗 AI Companions - xAI’s new Grok-powered interactive avatars

    🗣️ Voxtral - Mistral’s first open-source voice model for speech understanding

    📓 NotebookLM - New featured notebooks for expert advice on specific topics

    📚 Claude Directory - A new directory of tools that connect to Claude

💼 AI Job Opportunities

    🗂️ Deepmind - Administrative Business Partner

    💡 Cresta - Demo Engineer

    🤝 Databricks - Associate Account Executive

    🏗️ Glean - Solutions Architect

📰 Everything else in AI today

Mistral unveiled Voxtral, a low-cost, open-source speech understanding model family that combines transcription with native Q&A capabilities.

Google revealed that its AI security agent, Big Sleep, discovered a critical security flaw that allowed Google to stop the vulnerability before it was exploited.

U.S. President Donald Trump announced over $92B in AI and energy investments at a Pennsylvania summit, saying America’s destiny is to be the “AI superpower.”

Google is investing $25B in data centers and AI infrastructure across the PJM electric grid region, including $3B to modernize Pennsylvania hydropower plants.

Anthropic launched Claude for Financial Services, a solution that integrates Claude with market data and enterprise platforms for financial institutions.

Nvidia plans to resume sales of its H20 AI chip to China after CEO Jensen Huang received assurances from U.S. leadership, with AMD also resuming sales in the region.

COMMUNITY

🎥 Join our next live workshop

Join our next workshop this Friday, July 18th, at 4 PM EST with Dr. Alvaro Cintas, The Rundown’s AI professor. By the end of the workshop, you'll be able to use Perplexity Comet (and other alternatives) to automate your browsing experience.

RSVP here. Not a member? Join The Rundown University on a 14-day free trial.

See you soon,

Rowan, Joey, Zach, Alvaro, and Jason—The Rundown’s editorial team

Fish AI Reader

Fish AI Reader

AI辅助创作,多种专业模板,深度分析,高质量内容生成。从观点提取到深度思考,FishAI为您提供全方位的创作支持。新版本引入自定义参数,让您的创作更加个性化和精准。

FishAI

FishAI

鱼阅,AI 时代的下一个智能信息助手,助你摆脱信息焦虑

联系邮箱 441953276@qq.com

相关标签

Thinking Machine Labs AI融资 多模态AI 开源AI Mira Murati
相关文章