FieldAI builds foundational AI models that help all kinds of robots learn and adapt to new environments using physics. https://t.co/keiLCWL3Zc ...

投资了 Pinterest、Shopify、Twilio、Yelp、LinkedIn 等公司的投资机构 Bessemer Venture Partners 发布了他们的《The State of AI 2025》，研究了 20 家高增 长、具备持久竞争力的 AI 创业公司，包括表现突出的 Perplexity、Abridge 和 Cursor 等，对于今年的 AI 创业、基模的进展、Infra 的进展以及未来创业趋 势进行了梳理和总结。 现状有好有坏。 对于未来，Bessemer Venture Partners 认为，浏览器的竞争会更卷、视频生成将在明年进入正式可用状态，以及评估（Evals）与数据溯源将成为 AI 产品开 发的刚需。 还提到了已经成为开源核心势力的中国 AI 厂商们，总之，作为一篇了解 2025 AI 创业现状的报告，很合适。 原文链接：https://www.bvp.com/atlas/the-state-of-ai-2025 注：本次报告用了很多超新星、流星、星系的类比（怎么看怎么奇怪），本来想去掉，但为了不影响一些关键观点的理解，就尽量保留了。 TLDR: AI 创业，Day One ...

Partnerships & Integrations - Cohere 与加拿大电信供应商 BCE 达成协议，将其 AI 模型整合到 BCE 提供的服务中 [1] AI Industry - 人工智能初创公司 Cohere 积极拓展其 AI 模型在不同行业的应用 [1]

Core Viewpoint - Meta has declined to sign the European Union's code of practice for its AI Act, citing concerns over legal uncertainties and regulatory overreach that could hinder AI development in Europe [1][2][3] Group 1: Meta's Position - Meta's chief global affairs officer, Joel Kaplan, criticized the EU's Code of Practice for general-purpose AI models, stating it introduces legal uncertainties and exceeds the AI Act's intended scope [2] - Kaplan described the EU's legislation as "over-reach," arguing it could stifle the development and deployment of advanced AI models in Europe and negatively impact European companies aiming to build businesses on these technologies [3] Group 2: EU's AI Regulations - The EU's code of practice is a voluntary framework designed to assist companies in complying with upcoming AI regulations, requiring regular updates and documentation about AI tools, banning training on pirated content, and adhering to content owners' requests [2] - The AI Act categorizes certain AI applications as "unacceptable risk," outright banning them, while defining "high-risk" uses that require registration and compliance with risk and quality management obligations [3] Group 3: Industry Response - Major tech companies, including Meta, have been opposing the EU's AI regulations and have urged the European Commission to postpone the implementation timeline, which the Commission has refused to alter [4] - New guidelines for AI model providers were published by the EU, with compliance deadlines set for August 2, 2027, for companies with general-purpose AI models that pose systemic risks [5]

Product & Technology - The company is developing an open-source, human-robot interaction platform named Richie, designed with emotional expression capabilities [1][2] - Richie is designed as a kit, similar to Lego, to encourage community tinkering and improvement through open-source contributions [3][4] - The platform includes sensors, cameras, microphones, and speakers, and is compatible with AI models for speech and image processing [2] - The company will provide an SDK to facilitate the creation of new robot behaviors and applications, encouraging community sharing and accessibility [6] Market & Business Model - The company aims to provide an accessible platform for building human-robot interactions, differentiating itself from expensive humanoid robots [2][7][9] - The open-source nature of Richie is intended to maximize the value of community-built behaviors and applications [9] Sales & Distribution - The company has already generated $1 million in revenue from this project within a few days of launch [11] - Approximately 60% of initial sales are from the U S market, with the remaining 40% from international markets, primarily Europe, Asia, and South America [11] Manufacturing & Sourcing - The company is exploring local sourcing and manufacturing options in various locations to leverage design simplicity and navigate trade environments [12][13]

RT s4mmy (@S4mmyEth)Microsoft has just unveiled that its AI models are performing medical diagnosis with >80% accuracy.While I don’t see this as a threat to replacing doctors, it’s a huge step forward in reducing the workload for frontline medical staff.Not to mention mitigating the potential indemnity claims for misdiagnosis.As someone who has several family members in the medical profession, these are frequently voiced as their biggest concerns.ASI will disrupt every industry from Medical, insurance and m ...

RT s4mmy (@S4mmyEth)Microsoft has just unveiled that its AI models are performing medical diagnosis with >80% accuracy.While I don’t see this as a threat to replacing doctors, it’s a huge step forward in reducing the workload for frontline medical staff.Not to mention mitigating the potential indemnity claims for misdiagnosis.As someone who has several family members in the medical profession, these are frequently voiced as their biggest concerns.ASI will disrupt every industry from Medical, insurance and m ...

Core Viewpoint - A federal judge ruled in favor of Meta in a lawsuit regarding the training of AI models on copyrighted works, determining that such use falls under the "fair use" doctrine of copyright law [2][3]. Group 1: Legal Rulings - Judge Vince Chhabria issued a summary judgment, allowing Meta to avoid a jury trial, and concluded that the training of AI models on copyrighted books was legal under fair use [2]. - The ruling is part of a broader trend favoring the tech industry, as seen in a similar case involving Anthropic, although both rulings are limited in scope [3]. - Judge Chhabria emphasized that this decision does not imply that all AI training on copyrighted materials is lawful, but rather that the plaintiffs failed to present adequate arguments and evidence [4][5]. Group 2: Market Impact and Evidence - The judge noted that the plaintiffs did not provide meaningful evidence to demonstrate that Meta's actions harmed the market for the authors' works, which is crucial in copyright cases [8]. - The ruling highlighted that the transformative nature of Meta's AI models, which do not simply reproduce the authors' books, played a significant role in the decision [5]. Group 3: Industry Context - The outcomes of the cases involving Meta and Anthropic are part of ongoing legal battles faced by technology companies regarding the use of copyrighted materials for AI training, with other lawsuits pending against companies like OpenAI and Midjourney [9]. - Judge Chhabria pointed out that fair use defenses are highly case-specific, suggesting that different industries may have varying strengths in their fair use arguments [10].

Company Overview - Rubrik, a cybersecurity software company backed by Microsoft, is acquiring Predibase, a startup focused on AI model deployment [1][3] - The acquisition is part of Rubrik's strategy to enter the rapidly growing AI market [3][4] - Rubrik has achieved over $1 billion in annualized revenue by providing data backup solutions for clients [4][7] Acquisition Details - The financial terms of the deal are not publicly disclosed, but estimates suggest Rubrik will pay between $100 million and $500 million for Predibase [2] - Predibase will operate as a separate unit post-acquisition, continuing to support customers in connecting data from various third-party systems [6] Market Context - The AI market is experiencing significant growth, with cloud infrastructure providers generating billions from services related to AI model training [3] - Companies like Anthropic and OpenAI are rapidly expanding by offering subscription-based chatbot services, indicating a strong demand for AI tools [3] Predibase Background - Founded in 2021, Predibase has received over $28 million in investments from venture firms and has notable clients including Checkr, Marsh McLennan, and Qualcomm [5] - The startup employs 25 people and was co-founded by individuals with prior experience in AI tools at Uber [5]

Google, xAI, and other tech companies aren't wasting any time distancing themselves from Scale AI. Just hours after Meta announced a massive $14.3 billion investment in the data labeling startup, Google halted multiple projects with the company, according to internal correspondence obtained by Business Insider. Google had already been winding down its work with Scale AI, two contractors working on its projects told BI.OpenAI and Elon Musk's xAI have paused some Scale AI projects, too, Scale AI contractors ...