• The Recap AI
  • Posts
  • Ive & OpenAI: A $6.5B vision for AI hardware

Ive & OpenAI: A $6.5B vision for AI hardware

PLUS: Google AI Search evolves, Mistral's new Devstral model, and Nvidia on China chip policy

Good morning, AI enthusiast.

A landmark $6.5 billion deal sees OpenAI team up with Jony Ive, the renowned designer from Apple's golden era. Their goal is to forge new AI-centric hardware through the acquisition of Ive's startup, 'io', potentially reshaping our interaction with technology.

The collaboration hints at a future where we interact with AI beyond current screens and keyboards. Could this be the dawn of a new era in personal technology, potentially shifting the paradigm away from smartphones as we know them?

In today’s AI recap:

  • OpenAI’s $6.5B AI hardware vision with Jony Ive

  • Google expands AI Mode in Search for US users

  • Mistral’s new Devstral model for developers

  • Nvidia CEO on ‘failed’ US chip curbs on China

Ive's Big Bet: OpenAI Acquires Hardware Startup

The Recap: Jony Ive, the design visionary behind many Apple icons, is joining OpenAI to spearhead the creation of new AI-centric hardware. This move comes via OpenAI's reported $6.5B deal to acquire Ive's startup 'io', signaling a bold step towards AI-native devices.

Unpacked:

  • Ive will assume deep design responsibilities across OpenAI and 'io', aiming to 'completely re-imagine what it means to use a computer' with AI at its core.

  • OpenAI purchased the remaining stake in 'io', integrating its team; Ive contributes creatively while his design firm LoveFrom stays independent.

  • CEO Sam Altman states these devices will offer a radically improved interaction model for AI, potentially moving beyond current screen-and-keyboard paradigms and challenging smartphone dominance.

Bottom line: OpenAI is clearly betting big on dedicated hardware to unlock AI's full potential, potentially shaping the next wave of personal technology. Professionals should anticipate new tools that could fundamentally alter how they interact with AI and boost productivity.

Google Search Goes All-In on AI

The Recap: Google significantly expands its AI Mode in Search, making the advanced feature, announced at I/O 2025, accessible to all users in the US. This move marks a fundamental shift in how Google presents information, moving beyond traditional link-based results.

Unpacked:

  • AI Mode, previously a Labs experiment, now gives everyone on Google Search the option to engage with AI for answers.

  • Google enhances AI Mode by allowing personal data integration from services like Gmail to tailor responses, while providing users control over this feature.

  • The company also previews Deep Search, an upcoming agentic capability that will gather extensive data to generate comprehensive reports for complex queries. You can learn more about its integration with Gemini.

Bottom line: This widespread rollout of AI-powered search signals a major transformation in how users will find and interact with information online. Expect search to become more conversational and results more synthesized, as Google pushes towards increasingly intelligent systems.

Mistral Unleashes Devstral for Coders

The Recap: Mistral AI just launched its new Devstral model, an open-source LLM specifically designed to help developers tackle complex software engineering tasks by understanding and working with large codebases.

Unpacked:

  • Devstral achieves 46.8% on SWE-Bench Verified, outperforming previous open-source models by tackling real-world GitHub issues, moving beyond simple code completion.

  • Mistral AI developed Devstral in collaboration with All Hands AI, focusing on agentic capabilities that allow the model to navigate and modify intricate code structures effectively.

  • Developers can access the model for local deployment on hardware like a single RTX 4090 or a Mac with 32GB RAM, or use it via Mistral's API platform, la Plateforme.

Bottom line: Devstral's release provides developers with a powerful, open-source tool that can significantly enhance coding workflows and autonomous software development. Its strong benchmark performance and accessibility signal a practical step forward for AI-assisted engineering.

Nvidia CEO: US Chip Curbs on China 'Failed'

The Recap: Nvidia CEO Jensen Huang bluntly stated that U.S. chip export controls targeting China have been a "failure." He argued the policies significantly cut Nvidia's market share in China while simultaneously accelerating Beijing's drive to develop its own domestic chip capabilities.

Unpacked:

  • Nvidia's dominance in China plummeted from nearly 95% to 50%, directly impacting revenue due to restrictions, such as those affecting its H20 chips.

  • The U.S. export controls, aimed at limiting China's access to advanced AI technology as outlined in policies like the General Prohibition 10 Guidance, inadvertently spurred local Chinese firms to innovate faster.

  • The policy impact is evident as Chinese tech giants like Huawei advance their own chip capabilities, complicating the US-China tech dynamic, with Huang noting China is 'right behind' the U.S. in AI.

Bottom line: Huang's candid assessment highlights the intricate challenge of balancing national security interests with technological leadership and global market access. This situation underscores the rapid evolution of the global AI landscape, where policy decisions can significantly reshape competitive advantages and innovation trajectories.

The Shortlist

Meta launched its Llama Startup Program, offering resources, technical support, and potential cloud cost reimbursements to early-stage startups in the US building generative AI applications with Llama models.

Microsoft suffered an accidental leak during its Build conference when its AI security head's screen share revealed Walmart's AI expansion plans, including use of 'Entra Web and AI Gateway,' via a visible Teams message.

Jailbreaking exploits vulnerabilities in most leading AI chatbots, enabling them to generate dangerous or illegal information according to a new study from Ben Gurion University researchers highlighting the threat of 'Dark LLMs'.

Uber Freight unveiled a suite of new AI-powered features, including Insights AI for data analysis and over 30 AI agents to automate logistics tasks, aiming to enhance supply chain efficiency for shippers like Colgate-Palmolive.

What did you think of today's email?

Before you go we’d love to know what you thought of today's newsletter. We read every single message to help improve The Recap experience.

Login or Subscribe to participate in polls.

Signing off,

David, Lucas, Mitchell — The Recap editorial team