Apple’s next CEO to enter the AI war

In Today’s Edition,

• AI has blind spots. Cuey cross-checks models to find them

• Apple Names John Ternus as Next CEO

• Amazon and Anthropic Expand AI Partnership

• Stock Updates

• Anthropic introduces live Artifacts in Cowork

• AI Highlights of the Week

• Too Important to Miss

AI has blind spots. Cuey cross-checks models to find them

AI is confidently wrong too often. If you're using it for research or decisions, trusting one answer is a real risk. Why let one model make the whole call?

Cuey gives you multiple perspectives by cross-checking ChatGPT, Claude, and Gemini side-by-side in one tab. Compare answers instantly to get the best result.

• Catch hallucinations before you act on them

• Choose from 30+ models

• Your chat history travels model-to-model

Apple Names John Ternus as Next CEO

Apple confirms a planned leadership transition effective September 1, 2026

Apple says Tim Cook will move from CEO to Executive Chairman of the board, while John Ternus (SVP, Hardware Engineering) becomes the next Apple CEO on September 1, 2026, following a unanimous board approval.

Key Points:

• Leadership change is set for Sept 1, 2026: Cook transitions to Executive Chairman; Ternus becomes CEO.

• Cook stays CEO through the summer to support a “smooth transition,” and will continue engaging with policymakers in his new role.

• Board leadership reshuffle: Current non-executive chairman Arthur Levinson becomes lead independent director, and Ternus joins Apple’s board on Sept 1, 2026.

• Apple frames Ternus as a long-tenured product leader: He joined Apple in 2001, became a VP in 2013, and joined the executive team in 2021, overseeing hardware across major product categories.

• Apple highlights Cook-era scale: The release notes growth in market cap and revenue under Cook’s tenure, plus expansion of Apple’s footprint and installed base.

Why It Matters:
This is a major planned succession at the top of Apple, shifting day-to-day leadership to a longtime hardware executive, while keeping Cook closely involved as Executive Chairman during a period when Apple’s product roadmap, services expansion, and global policy pressure are all high-stakes priorities.

Amazon and Anthropic Expand AI Partnership

Anthropic secures massive Trainium capacity as Amazon expands funding and brings Claude’s full platform into AWS

Amazon and Anthropic are deepening their partnership, with Anthropic set to spend over $100 billion on AWS in the next decade and Amazon investing $5 billion now, with up to $20 billion more tied to milestones.

Key Points:

• 5 gigawatts of AI capacity: Anthropic will secure up to 5GW of current and future AWS capacity to train and run its models, including meaningful Trainium3 capacity expected to come online this year.

• Trainium becomes the backbone: The commitment spans Trainium2, Trainium3, Trainium4, plus options to buy future Trainium generations as they launch, aimed at better price-performance for training and inference.

• Claude Platform inside AWS: AWS customers can access the full Anthropic-native Claude console directly within AWS using their existing AWS accounts, without separate credentials, contracts, or billing setup.

• Amazon invests more: Amazon will invest $5B today and may invest up to $20B more later based on commercial milestones, on top of the $8B it previously invested in Anthropic.

• Project Rainier scale: The companies highlighted Project Rainier, described as one of the world’s largest AI compute clusters, with nearly half a million Trainium2 chips, used to train and deploy Claude models globally.

• Global expansion for inference: The partnership includes a “meaningful expansion” of inference capacity in Asia and Europe to better serve Claude’s international user base.

Why It Matters:
The deal highlights a major shift toward custom chips and hyperscale infrastructure as key advantages in frontier AI. It gives Anthropic long-term compute for Claude, while strengthening AWS and pushing Trainium deeper into AI training and inference.

Stock Updates

Anthropic introduces live Artifacts in Cowork

AI Highlights of the Week

• Gemini wants to catch Claude

  Google has formed a dedicated team to improve Gemini’s coding skills, especially for complex software tasks.

  The push comes as Google sees Anthropic’s coding tools as stronger, driving more focus on internal code training and AI development.

• NSA Gets Access to Anthropic’s Mythos
  The NSA reportedly has access to Anthropic’s restricted Mythos model even as the Pentagon still labels the company a supply-chain risk.

  The contradiction exposes a growing split inside Defense, with Anthropic’s AI apparently staying in use for cybersecurity work even while contractors are told to steer clear.

• Meta Tracks Employee Keystrokes for AI Model
  Meta plans to use employee keystrokes and mouse movements as training data to build AI models that better understand how people use computers.

  The move gives Meta a new source of real-world interaction data, but also raises fresh concerns around privacy and how far AI companies will go to train their systems.

• Google Eyes New AI Chip Deal With Marvell

  Google is in talks with Marvell to build custom AI inference chips, expanding beyond Broadcom as it scales its AI infrastructure.

  The move signals a shift toward inference-focused compute, where costs now dominate, and highlights Google’s push to diversify its AI chip supply chain for better efficiency and scale.

Too Important to Miss

Blue Origin’s New Glenn Misfires on Customer Launch

The Real AI Data Center Bottleneck Is Skilled Labor

Google Adds AI Studio to Pro and Ultra Plans

Google Brings AI Mode Side by Side Into Chrome

Feedback

Know someone who may be interested?