Is Copilot Microsoft's Everything App?

The episode starts with host Brian McCulloch introducing the main tech news of the day, including Microsoft's unveiling of its co-pilot super app, updates on Tesla's Optimus robot production, Meta's new AI game-building apps for Instagram and Facebook, and a rapidly valued startup, TypeSafe, which shifted from a $200 million valuation to over $10 billion within a week. The host also mentions a new online card game designed for IT professionals by Eaton, called Ace of Uptime, with new decks tailored to enterprise and data center environments. The episode sets the stage for a deep dive into Microsoft's new co-pilot app, updates on Tesla's Optimus production challenges, Meta's push into AI-generated gaming content, and the dramatic valuation leap of startup TypeSafe. Additionally, the episode highlights the significant AI infrastructure investments projected to reach over $10 trillion in the U.S., emphasizing the massive economic implications of AI development.

Microsoft officially launched its new co-pilot 'super app,' integrating chat, coding, and agents into a unified interface, and rebranded its AI assistant Scout as Autopilot. The app features three main tabs: home, code, and autopilot. The home tab provides an overview with activity review, suggestions, and a personalized dashboard. The code tab, a significant addition, allows users to create apps, trackers, or automation by describing them in plain language, leveraging the same tech as GitHub Copilot, for internal use within Microsoft 365. Autopilot is presented as a digital teammate, capable of running in the background to monitor channels, handle follow-ups, and execute recurring tasks. It is built with enterprise security in mind, residing within the organization’s tenant and functioning with its own identity, memory, and workspace. The app's usage-based billing model means companies will pay based on how much the AI features are used, managed via Microsoft’s new AI cost management tools. This restructuring reflects Microsoft's effort to settle into a consistent AI assistant formula, merging consumer and commercial features and establishing a single point of control.

Tesla has increased Optimus robot production to several hundred units weekly, a significant ramp-up from a few dozen in Q2, but faces issues with robotic hands, automation equipment, and supplier constraints. The company aims to establish a continuous automated line capable of over 1,000 robots per week by year-end, although this is still short of the planned 20,000 units weekly. Tesla is still refining the V3 version of Optimus, which is lighter, more camera-equipped, and designed to look more human, but it does not yet meet durability and reliability standards. Most robots are currently used internally for testing and data collection in Tesla’s controlled environments. Production bottlenecks include difficulties in assembling the robot’s hands, which contain over 100 tiny components, and issues with durability and sensor reliability. Elon Musk has indicated plans to sell Optimus at $20,000 to $30,000, but Tesla remains in the development and testing phase, addressing manufacturing complexity and cost challenges.

Meta has announced Horizon Create, a mobile app, and Horizon Studio, a web app, both aimed at enabling users to build games with AI-generated prompts. These tools, which will be available in early access, will support games on Facebook, Instagram, and Horizon. Meta’s strategy appears to be to improve Horizon’s reach by promoting games made with these tools directly within its social platforms, where users can discover, play, and share games quickly, without downloading separate apps. The move follows similar efforts by Roblox to incorporate AI into game creation, though skepticism remains about the engagement levels of AI-generated games compared to handcrafted ones. Meta’s approach is to embed these games naturally into existing feeds and social experiences, hoping that engaging, stable, and culturally relevant content will increase their popularity and reach.

Sources report that TypeSafe, a developer of AI models called Jeff, is in advanced talks to raise over $1 billion at a valuation exceeding $10 billion, despite only having announced a seed round just last week at a $200 million valuation. This rapid increase exemplifies what some characterize as the fastest startup valuation growth in history. High-profile investors and industry observers are praising the company's potential, with notable mentions from Nvidia's Dion Harris and other AI leaders. Similarly, OpenEvidence, an AI search engine for physicians, recently raised $250 million at a $15 billion valuation, marking a significant jump from $12 billion in January. This rapid funding growth indicates a booming investor interest in healthcare AI and alternative AI architectures, especially as large companies and investors race to capitalize on the immense economic opportunities in AI infrastructure.

A long read highlights that AI infrastructure investment in the U.S. is on track to become the largest speculative capital buildout in history. Economists project a total of $10.3 trillion will be invested in data centers and AI-related infrastructure from 2025 to 2032, which would constitute about 3.6% of the U.S. GDP annually. This massive investment surpasses historical infrastructure booms like railroads and highways. Goldman Sachs also estimates that AI investment will reach about 1.9% of the U.S. GDP in 2026. This scale of investment underscores how deeply AI is becoming embedded in the American economy, potentially reshaping virtually every industry and emphasizing the strategic importance of AI infrastructure.