AI Pulse

AI summaryDuring Amazon's earnings call, CEO Andy Jassy addressed concerns about external AI shopping agents. He stated that these agents are "still not great, nor more convenient than coming to a retailer directly." Jassy reassured investors that they should not be concerned about bots potentially disrupting Amazon's online store or rendering it obsolete.

AI summaryNvidia is reportedly in discussions to offer Nscale, an emerging AI cloud provider, billions of dollars in credit guarantees. This financial backing would support Nscale's development of a 460MW data center. The data center is intended for use by Anthropic, indicating a strategic collaboration within the AI infrastructure sector.

AI summaryMicrosoft's cloud business is experiencing accelerated growth, with first-quarter revenue reaching $92.3 billion, a 19% year-over-year increase, surpassing Wall Street analyst expectations. The company's profit also exceeded predictions, totaling $36.6 billion. This growth comes as Microsoft continues to invest heavily in building out its AI business.

AI summaryAlphabet's Google Cloud revenue more than doubled to $30.7 billion in the third quarter, partially offsetting slower growth in Google Search and YouTube. Despite this, investors remain concerned about the sluggish 12% growth in Google Search, which fell short of the 15% analyst expectation and marked a slowdown from 17% and 19% growth in previous quarters.

AI summaryWorkhorse has partnered with UK-based RHAELM, entering the modular AI infrastructure market. Under a Spec Unit Agreement, RHAELM will pay Workhorse $10 million to build a full-scale TerraPod manufacturing spec unit at Workhorse's Union City, Indiana facility. A non-binding Letter of Intent also outlines a path to a definitive Supply Agreement, targeting execution by December 31, 2026, which is framed to include a first order provision for 150 TerraPod systems.

AI summaryThrive Holdings, backed by OpenAI, has acquired physical security company ADT for $7.6 billion in equity, with a total enterprise value of $15.3 billion. Thrive Holdings, a spinout from Thrive Capital, aims to revitalize acquired companies with AI, with OpenAI supplying AI models and engineers. ADT will form the foundation of Thrive Holdings’ smart home security business, with plans to invest in and release new AI-powered security products. This acquisition comes as AI-powered physical security gains traction, with other tech giants also entering the field.

AI summaryIt is becoming increasingly difficult to secure funding for AI datacenters, signaling a potential AI bubble. Reports indicate that Wall Street's AI debt machine is starting to face supply issues, making it harder to raise capital. This trend suggests a future consolidation among the numerous new cloud providers, with many expected to struggle or fail as the market matures and funding tightens.

AI summaryFormer Palantir executives Andy Elder and Arjun Mathur, Stanford computer science alumni, have launched a new startup called Dream. The company secured $17 million in funding from a16z to develop AI world models specifically for the construction industry. Elder previously led product design for defense software at Palantir, while Mathur was a senior vice president in their commercial practice.

AI summaryAn OpenAI user reported three unauthorized charge attempts from "OpenAI OpCo LLC" on their credit card, despite having no active subscriptions or API credit purchases since September 2025. The charges were rejected because the user had previously blocked the card. The user's account is secured with a Google account, which has not been breached, leading them to question OpenAI's practices.

AI summaryWabi, an AI app now calling itself a “personal agent,” has successfully raised $55 million in its latest funding round, achieving a $1.25 billion valuation. This investment highlights the ongoing competition among investors to back leading AI assistants. Notable participants in the round included Andreessen Horowitz (a16z), Forerunner, Amazon Alexa Fund, and several prominent angel investors such as Shopify CEO Tobi Lütke and Salesforce CEO Marc Benioff.

AI summaryAnthropic's S-1 IPO filing is being analyzed by Joey Brookhart of SemiAnalysis, revealing a narrow market despite $120 billion ARR between leading AI labs. The top 1% of customers spend $50,000-$100,000 per employee, while broader Fortune 500 or Global 2000 companies show less IT spend. Significant token throughput gains, five to ten times higher year-over-year, and continued compute efficiencies are anticipated, addressing hardware and software interactions.

AI summaryAnthropic's S-1 filing reveals significant financial commitments, totaling hundreds of billions of dollars, primarily for compute and infrastructure capacity. These non-cancelable purchase commitments and lease agreements, with terms ranging from one to 20 years, suggest a strategy focused on massive investment in computational resources. The company's business plan appears to prioritize extensive spending on compute, with the justification that such investment is crucial for its survival and long-term goals.

AI summaryOpenAI is reportedly planning to purchase Google’s Tensor Processing Units (TPUs) to enhance its AI computing capacity and lessen its dependence on Nvidia. This move comes as Amazon, Microsoft, and Alphabet collectively plan to invest up to $600 billion in AI infrastructure this year, underscoring the significant capital needed to address the increasing demand for AI computing. The financial details of the deal remain undisclosed, and neither company has confirmed the reported acquisition.

AI summaryMicrosoft reported $10 billion in revenue from commercial arrangements with OpenAI in the quarter ended September 30. This figure represents nearly 11% of Microsoft’s total quarterly revenue of $92.3 billion, underscoring the significant financial relationship between the two companies. OpenAI reportedly shares 20% of its revenue with Microsoft as part of a previous deal, and this revenue-sharing agreement is set to continue through 2030, subject to a cap.

AI summaryAnthropic has launched tools to help companies understand their AI spending, following a successful period where its AI coding tool, Claude Code, achieved significant business success. The company reported an operating profit in the second quarter and a 900% net revenue retention from Q3 2025 to Q3 2026. Despite perceptions of heated competition, particularly with OpenAI's Astra, Anthropic's tools have gained traction due to strong enterprise engagement and effective standardization of workflows, making switching to competitors less appealing for many users.

AI summaryOpenAI and ServiceNow are expanding their partnership to integrate more models and agents into ServiceNow products, connecting workflows with ChatGPT and Codex. This collaboration aims to simplify the development of applications, resolution of IT requests, and addressing of security issues using AI. The ServiceNow MCP plugin is currently available, with enhanced capabilities expected by late 2026. Voice capabilities in the Super Intelligence Control Tower, OpenAI-powered security features, and Codex in AI Workflow Factory are all anticipated for 2027.

AI summaryOpenAI and Oracle are expanding their partnership, integrating Oracle applications and data into ChatGPT and leveraging OpenAI models for new Oracle products. This collaboration aims to assist businesses in utilizing AI for core operations, application development, and enhanced security. Building on their existing commercial relationship, Oracle customers in the U.S. can now access OpenAI API, Codex, and ChatGPT Work via Oracle Marketplace using Universal Credits.

AI summaryAnthropic, valued at over $1.5 trillion in its IPO, has generated significant wealth for its cofounders. The company acknowledges potential risks associated with its technology, including the erosion of trust, societal harm, geopolitical destabilization, and even catastrophic or existential risks to humanity. These details were further elaborated in the IPO prospectus.

AI summaryOpenAI's FDE (Full-Stack Deep Learning Engineers) team addresses the challenge of rapidly evolving AI models like Codex. They focus on building platforms and defining model requirements for customers, while the Research team closes capability gaps. An August 2026 example showed an initial deployment taking two months, but extending the solution to a second team took only one day, demonstrating accelerated scaling.

AI summaryOpenAI and Google Cloud are partnering to integrate OpenAI's frontier models and Codex agentic workflows into Google Cloud's Model Garden. This collaboration will provide customers with native access to these models, enabling complex agentic workflows at scale, supported by Google Cloud’s data governance and high-performance infrastructure. Additionally, GPT models will soon be available in Antigravity, Google’s agent-powered coding tool. OpenAI models will run entirely on Google’s infrastructure, with transactions managed through the Google Cloud Marketplace, making it easier for businesses with existing Google Cloud spending commitments to adopt GPT models.

AI summaryOpenAI developed a novel ad stack for ChatGPT, differing from traditional systems by adapting to evolving user intent across conversations. This bespoke system uses a custom, post-trained frontier LLM with a tool call architecture for retrieval, ranking, relevance, and auctioning. Key design decisions include stateful ad sessions, LLM-driven context interpretation for ad delivery, and separating ad selection from answer generation to maintain neutrality. The LLM orchestrates the ad flow, determining optimal ad surfacing based on conversation history and current turns.

AI summaryOpenAI has acquired Orchestra Bio, a platform designed to assist biopharma teams in coordinating the intricate drug development process. This acquisition integrates Orchestra's team and technology with Rosalind, OpenAI's AI models and tools for life sciences research. The aim is to empower scientists to accelerate discoveries and streamline the management of turning these into treatments, ultimately reaching patients sooner. Developing new medicines typically takes 10 to 15 years, and AI is now at an inflection point to help with these complex workflows.