The afternoon briefing.
SpaceX plans a $55B AI chip plant, OpenAI unveils new voice models with GPT-5 level reasoning, and Apple's camera-equipped AirPods near production.
AI Infrastructure. The race for AI compute intensifies as SpaceX announces a colossal $55 billion plan to build an AI chip plant in Texas, with potential for even greater investment. This move underscores the immense capital flowing into foundational AI hardware. Meanwhile, CoreWeave reported its revenue more than doubled in the first quarter, highlighting the rapid growth in demand for specialized AI cloud services. These investments are critical for scaling future AI capabilities.
Model Advancements and Safety. OpenAI has unveiled new voice models, including GPT-Realtime-2, which offers GPT-5-level reasoning for real-time conversations and translation across 70 languages. Concurrently, the company introduced a 'Trusted Contact' safeguard for ChatGPT users, allowing emergency contacts to be notified in cases of potential self-harm. These developments showcase both the rapid progression of AI capabilities and the growing focus on responsible deployment and user safety.
On-Device AI and Agents. The push for AI integration into everyday devices continues, with Apple's camera-equipped AirPods reportedly nearing mass production, suggesting a future of ambient, on-device AI. Perplexity's Personal Computer, an AI agent experience, is now broadly available for Mac users, expanding access to personalized AI assistance. Furthermore, AWS has launched agentic AI payment capabilities, indicating a move towards autonomous AI systems handling financial transactions.
Regulatory Landscape. The regulatory environment for AI remains fragmented and evolving. The Golden Globes adopted AI rules less strict than the Oscars', illustrating differing industry approaches to generative AI. In Europe, the implementation of the EU AI Act faces delays, with deadlines for high-risk AI pushed back to late 2027 or 2028. This suggests a cautious, phased approach to managing the complexities of comprehensive AI governance.
SpaceX plans $55 billion AI chip plant in Texas, could reach $119 billion
Elon Musk's SpaceX is planning to invest at least $55 billion into its "Terafab" chip plant in Austin, Texas. Public hearing notices indicate the investment could balloon to $119 billion if additional phases are constructed.
CoreWeave revenue more than doubles in first quarter, topping estimates
CoreWeave, an AI cloud provider, reported its revenue more than doubled in the first quarter, exceeding analyst estimates. The company has been raising billions in debt to finance its data center expansion for leading AI firms.
OpenAI introduces new ‘Trusted Contact’ safeguard for possible self-harm
OpenAI is expanding efforts to protect ChatGPT users by introducing a 'Trusted Contact' safeguard. This feature allows users to designate an emergency contact to be notified if conversations turn to self-harm.
Razer partners with 'P2P for AI' network for low-cost image generation
Razer utilized a decentralized GPU marketplace to generate over 11,000 AI images at just $0.01 per generation. This partnership achieved low costs and stable performance without relying on traditional cloud infrastructure.
Tesla Model Y is first car to meet new U.S. driver assistance safety benchmark
The 2026 Tesla Model Y, assembled on or after November 12, 2025, is the first vehicle to meet a new U.S. driver assistance safety benchmark. This rating applies to its advanced driver assistance systems.
The Golden Globes’ AI rules are way less strict than the Oscars’
The Golden Globes have released new guidelines for AI usage that are notably less stringent than those recently announced by the Academy Awards. This highlights a divergence in how major awards bodies are approaching AI.
Perplexity’s Personal Computer is now available to everyone on Mac
Perplexity's Personal Computer, which brings AI agents to the Mac platform, is now open to all users. This expands access to its AI-powered assistant capabilities for Mac owners.
Apple’s AirPods with cameras for AI are apparently close to production
Apple's rumored AirPods with cameras are reportedly nearing early mass production testing, according to Bloomberg. Prototypes are actively being used by testers in the design validation stage.
Mozilla says 271 vulnerabilities found by Mythos AI have almost no false positives
Mozilla reported that its use of the Claude Mythos preview AI found 271 vulnerabilities in Firefox with a high degree of accuracy. This demonstrates AI's potential for effective security bug detection.
OpenAI's new voice model brings GPT-5-level reasoning to real-time conversations
OpenAI is shipping three new voice models, including GPT-Realtime-2, which offers reasoning capabilities matching GPT-5. These models can also translate across 70+ languages and transcribe live speech.
Paul Tudor Jones says U.S. is late to regulating AI, should have already done it
Billionaire investor Paul Tudor Jones stated that the U.S. is behind on AI regulation and should have already implemented measures. He emphasized the heated rivalry with China for AI innovation leadership.
SAP completes acquisition of Reltio to make enterprise data AI-ready
SAP has completed its acquisition of Reltio, a move aimed at helping customers prepare their SAP and non-SAP enterprise data for AI applications. The acquisition strengthens SAP's data management capabilities for AI.
Europe delays most AI Act deadlines, eases requirements for small and medium businesses
The EU has agreed to simplified AI rules, pushing back deadlines for high-risk AI to late 2027 or 2028 and easing requirements for SMBs. Labeling for deepfakes and AI-generated text still takes effect in August 2026.
AWS launches agentic AI payment capabilities
AWS has introduced new agentic AI payment capabilities, marking the first purpose-built payment functionality for autonomous agents. This update aims to enable AI systems to handle payment processes more independently.