Find out why 100K+ engineers read The Code twice a week

Staying behind on tech trends can be a career killer.

But let’s face it, no one has hours to spare every week trying to stay updated.

That’s why over 100,000 engineers at companies like Google, Meta, and Apple read The Code twice a week.

Here’s why it works:

• No fluff, just signal – Learn the most important tech news delivered in just two short emails.

• Supercharge your skills – Get access to top research papers and resources that give you an edge in the industry.

• See the future first – Discover what’s next before it hits the mainstream, so you can lead, not follow.

Join 100,000+ engineers who read The Code to stay ahead of the curve.

Today:

• OpenAI Teams with Broadcom to Build 10GW of Custom AI Hardware

• Nano Banana Arrives in Google Search, NotebookLM, and Soon Photos

• Microsoft Launches MAI-Image-1—Now Top 10 on LMArena

• Young Australians Are Using AI to Work Smarter, Not Just Faster

• Google Gives Students Free Access to Gemini, Veo 3, and More

OpenAI goes all in on compute...

OpenAI and Broadcom have partnered to create and deploy 10 gigawatts of custom AI chips, a move that signals OpenAI’s entry into hardware design and a possible challenge to NVIDIA’s dominance. These chips, built specifically for OpenAI’s massive AI training and inference workloads, are part of a full-stack system co-designed with Broadcom. 

Deployment begins in late 2026. Sam Altman and Greg Brockman emphasized how this hardware effort will optimize performance, reduce costs, and support OpenAI’s push toward superintelligence. The chips are already showing early success by using AI itself to optimize their design. 

However, with financial structures involving merchant financing and overlapping deals with NVIDIA and AMD, some analysts warn of a potential AI bubble. Former Intel CEO confirms we're in a hype-driven era but notes the bubble may persist as real breakthroughs emerge. Despite financial concerns, this scale of investment could transform global AI infrastructure for years to come.

WATCH THE VIDEO ON YOUTUBE

Nano Banana Arrives in Google Search, NotebookLM, and Soon Photos

Google is expanding access to its powerful image model, Nano Banana, across Search, NotebookLM, and soon Photos. Users can now create and edit AI images directly in the Google app and enjoy enhanced video overviews, new visual styles, and quicker insights in NotebookLM.

KEY POINTS

• Nano Banana in Google Search: Use Lens + Create mode to instantly edit or generate images from photos.

• NotebookLM Upgrades: Adds 6 new visual styles (e.g., watercolor, anime), contextual illustrations, and a fast “Brief” format.

• Coming Soon to Photos: Nano Banana integration will enable creative editing directly in Google Photos.

Why it matters
This update puts powerful image creation tools right into everyday Google apps, making it easier for everyone to visualize ideas, learn faster, and express creatively. It turns AI from a standalone tool into a native feature of search, note-taking, and photo editing.

Microsoft Launches MAI-Image-1—Now Top 10 on LMArena

Microsoft AI has launched MAI-Image-1, its first in-house image generation model. Ranked in the top 10 on LMArena, it offers fast, photorealistic outputs with diverse styles. Designed with real creative needs in mind, it’s set to roll out in Copilot and Bing Image Creator.

KEY POINTS

• Top-Ranked Debut: MAI-Image-1 launches in the top 10 on LMArena, showcasing strong performance in the text-to-image field.

• Speed + Realism: Prioritizes photorealism, lighting, and visual diversity while remaining fast and flexible for creators.

• Product Integration: Will soon be available in Microsoft Copilot and Bing Image Creator for wider access.

Why it matters
This marks a major step in Microsoft’s push to own its creative AI stack. MAI-Image-1 brings fast, realistic image generation to tools people already use. That means more helpful, creative AI experiences directly in your daily workflow—designed with human needs and visual quality in mind.

Google Offers Top AI Tools Free to University Students in EMEA

Google is offering its top AI tools—Gemini 2.5 Pro, Nano Banana, Veo 3, and more—free to university students (18+) in Europe, the Middle East, and Africa for one year. The goal is to help students learn, think creatively, and build future-ready AI skills.

KEY POINTS

• Free AI Pro Plan for Students: Includes Gemini 2.5 Pro, Deep Research, Nano Banana, Veo 3, and 2 TB of storage for 12 months.

• Guided Learning Mode: Helps students with essays, math, quizzes, test prep, and step-by-step tutoring.

• Creative AI Tools: Students can design visuals, create video clips, and prototype projects using Nano Banana and Veo 3.

Why it matters

By giving students free access to powerful AI tools, Google is preparing the next generation for a tech-driven world. This program encourages deeper learning, creativity, and AI literacy—especially important as AI becomes a core skill in education, work, and daily life.

🧠RESEARCH

D2E: Scaling Vision-Action Pretraining on Desktop Data for Transfer to Embodied AI

The D2E framework trains AI using desktop interactions—like gaming—instead of costly real-world robot data. It compresses and labels gameplay data, then transfers the learning to physical tasks. D2E achieved up to 96.6% success in robotics benchmarks, proving desktop-based training can effectively prepare robots for real-world navigation and manipulation.

Thinking with Camera: A Unified Multimodal Model for Camera-Centric Understanding and Generation

Puffin is a new AI model that helps computers better understand and generate images from different camera views. It treats camera settings like a language, letting the model reason about space more effectively. Trained on 4 million examples, Puffin excels at tasks like scene generation, spatial reasoning, and photography guidance.

AutoPR: Let's Automate Your Academic Promotion!

AutoPR is a system that turns academic papers into engaging social media posts. Using a team of AI agents, it extracts key ideas, rewrites them for broader appeal, and tailors them for each platform. Compared to standard AI methods, AutoPR drastically boosts views, likes, and overall engagement with research.

🛠️TOP TOOLS

Each listing includes a hands-on tutorial so you can get started right away, whether you’re a beginner or a pro.

Wepik AI Slidesgo – AI Presentation Maker - browser‑based tool that turns a short prompt into a ready‑to‑edit slide deck

Toonify – AI Image Generator - web app that turns a face photo into stylized “cartoon”

Clipdrop AI – AI Image Generator & Editor - browser‑based and API‑driven toolkit for creating and editing visuals with AI.

📲SOCIAL MEDIA

🗞️MORE NEWS

• Gen Z Australians are embracing AI at work despite job fears. They’re teaching leaders, creating tools, and using AI to boost confidence, learn faster, and shape strategy—while remaining aware of its risks and limits.

• Slack is upgrading Slackbot into a full AI assistant that helps users plan, schedule, search, and summarize using natural language. It draws from workspace data and integrates with Outlook and Google Calendar for smarter collaboration.

• Salesforce launched Agentforce 360, a major upgrade to its AI agent platform with tools like Agent Script and Agentforce Builder. It integrates deeper with Slack and aims to lead enterprise AI amid rising competition.

• California passed the U.S.’s first law regulating AI chatbots, requiring safety safeguards and allowing lawsuits for harm. Governor Gavin Newsom signed it despite White House opposition, citing teen suicides linked to unregulated AI companions.

• Brookfield is investing up to $5 billion in Bloom Energy’s fuel cell tech to power AI data centers. The move supports cleaner energy, as demand for AI infrastructure grows and companies seek sustainable, scalable solutions.