Dec 11, 2025 6 min read Daily Rundown

🛎️ DeepSeek’s Forbidden GPUs

Plus: Local AI Coding, Google Finds Next Lens

Good Morning, AI Enthusiasts!

Nowadays, the AI story isn’t about models. It’s about who controls the compute, the tools, and the interface to reality itself.

COMPETITION

DeepSeek’s Blackwell Heist Shows Who Really Runs AI

👀 What’s happening: The Information says DeepSeek is training its next model on thousands of Nvidia Blackwell GPUs, which are chips the U.S. explicitly banned from China. Investigators call it smuggling, engineers call it survival. Nvidia denies wrongdoing, but the silence from everyone else is louder: no one builds frontier AI on wishful silicon.

🔥 How this hits reality: Let’s be honest. Without Nvidia, there is no “competitive” AI model. Google’s TPUs can hold the line for a bit, but the rest of the silicon scene might as well be decorative art installations. For teams, it means the moat isn’t data or algorithms anymore, it’s who can actually get GPUs through the border. As for the compliance deck, that’s not even the hard part.

🛎️ Our take: Forget “AI model sovereignty.” The frontier still runs on Nvidia chips.

TOGETHER WITH FELO

Stop Working FOR Your Documents, Make Them Work FOR You

The 10x knowledge worker isn't typing faster—they're directing AI agents while competitors drown in copy-paste hell. Right now, someone just translated a 50-page deck in 3 minutes while you're still reformatting slide 12. Felo LiveDoc is the world's first Agent Workspace where documents don't just sit there—they evolve, analyze, and update themselves. Imagine a phantom team of designers, translators, and analysts inside your docs, unifying scattered PDFs, videos, and spreadsheets into one intelligent system that grows smarter with every task.

Here's what makes you look like a genius: While legacy tools treat you like a glorified typist, LiveDoc's multi-agent system operates autonomously. Design agents polish your layouts. Data agents extract insights across PDFs, videos, and web pages simultaneously. Translation agents speak every language your customers do—all working 24/7 inside one living workspace. This isn't productivity theater—it's the death of document drudgery. You're not just saving time; you're unlocking a superpower your competitors don't know exists yet.

And it's not just you—it's your entire team: Think of your best team collaboration tool—now add tireless AI teammates who never sleep, never forget context, and execute instantly. LiveDoc expands collaboration from human-to-human to human-plus-AI: team members direct, AI agents execute, all on the same canvas in real time. No more "latest_final_v3" nightmares. While your competitors coordinate between people, you're coordinating an entire hybrid workforce.

The insider move? Early adopters are already 10x-ing their output while the rest of the market is still toggling between twelve tabs. Every day you wait is another day spent manually updating slides like it's 2015. The knowledge workers who win aren't the ones who work harder—they're the ones who stopped editing and started directing.

🚀 Join the revolution: Get early access to Felo LiveDoc and become the person everyone asks, "How did you finish that so fast?" The age of manual document labor just ended. Are you evolving, or still typing?

NEW LAUNCH

Mistral’s Devstral 2 Pushes Local AI Coding Mainstream

👀 What’s happening: Mistral unveiled Devstral 2, a new generation of code models built on a shared architecture: the 123B flagship for high-end agentic development, and a 24B “Small” version that runs offline on a single GPU or even a laptop. Both feature 256K context windows and top-tier SWE-Bench results, with the smaller model outperforming many 70B systems. Together with the Vibe CLI—a terminal-native interface for real-time project orchestration—Mistral’s lineup is built to live directly inside developers’ workflows.

🔥 How this hits reality: The release shifts AI coding from the cloud to the desk. Devstral Small 2’s local inference and edge efficiency open up AI-assisted programming for teams under strict data or latency constraints, while the full model pushes long-context reasoning closer to production scale. For enterprises, it means realistic on-prem AI coding; for independent developers, a self-contained, privacy-safe co-pilot that actually fits their hardware.

🛎️ Our take: Mistral just made frontier-grade coding intelligence portable, turning AI-assisted development from a service you call into a tool you own.

PREVIEW

Google Finds Its Next Lens After Gemini’s Win

👀 What’s happening: Google just previewed its Android XR smart glasses powered by Gemini, marking its first real comeback since the Glass collapse a decade ago. The two models — one voice-only, one with a built-in display — use Gemini’s multimodal brain to see, hear, and understand your surroundings. Users can ask it to translate a sign, recognize a landmark, or recall what they photographed earlier. Developer previews arrive in 2025, with a full commercial launch set for 2026.

🔥 How this hits reality: After Gemini 3 proved Google could finally build an AI that reasons rather than reacts, the company seems to have found the hardware form to match that brain. Android XR ties Gemini’s cognition directly to sensors, cameras, and real-world context, turning glasses into a live interface for spatial computing.

🛎️ Our take: Gemini gave Google intelligence; XR gives it embodiment. Together they hint at the first real successor to the smartphone.

QUICK HITS

Google launched a ₹399 (~$5) mid-tier AI Plus plan in India, offering expanded model access and 200GB storage to compete with ChatGPT Go.
Microsoft is investing CAD $19B to boost Canada’s AI and cloud infrastructure along with new sovereignty and talent programs.
Meta is reportedly going all-in on Avocado, a proprietary-leaning frontier model expected to arrive in 2026.
Inito raised $29M to use AI-designed antibodies to expand from fertility testing into a broader at-home health diagnostics platform.
Aetherflux aims to launch its first space-based data center in 2027, pitching orbital compute as a faster alternative to Earth’s energy-strained data centers.

TRENDING

Daily AI Launches

Incredible has launched an AI agent platform powered by Agent MAX, claiming 90% lower costs, 100x memory, and near-zero hallucinations.
SnapTodo has launched a visual planner combining a calendar, to-do list, and AI organization with a built-in booking system.
HERO has launched a workspace for structured docs (like contracts) that dynamically links files and data with database-like power.
Anytype has launched a private, secure collaboration hub for chatting, docs, and project management without data tracking.

Featured AI Tools

🧠 TRAE (The Real Al Engineer) is the most price-competitive coding agent.
🎥 KaraVideo unites all AI video models in one place.
📚 Heardly is the Fast Way to read Best Book.
🤖 Momen AI bridges the gap from vibe to viability, powering Lovable apps without code.

TOGETHER WITH US

AI Secret Media Group is the world’s #1 AI & Tech Newsletter Group, reaching over 2 million leaders across the global innovation ecosystem, from OpenAI, Anthropic, Google, and Microsoft to top AI labs, VCs, and fast-growing startups.

We operate the industry’s most influential portfolio of newsletters, each shaping a different frontier of the AI & Tech revolution:

Be Smarter in 5 Minutes