🤝 OpenAI & NVIDIA Strike 10 GW Partnership
OpenAI and NVIDIA plan to deploy 10 gigawatts of AI data centers powered by millions of NVIDIA GPUs. NVIDIA will invest up to $100B as each gigawatt comes online. The first 1GW site, using NVIDIA’s Vera Rubin platform, is set to launch in the second half of 2026.
📺 Gemini AI Lands on Google TVs
Google is bringing its Gemini AI assistant to Google TV, starting with TCL’s QM9K series. The AI offers natural language help for everything from show suggestions to school projects. Legacy voice commands remain, but Gemini adds smarter, conversational tools to over 300M devices.
🦾 UAE, NVIDIA Launch First AI & Robotics Lab
NVIDIA and Abu Dhabi’s TII have launched the first NVIDIA AI Technology Center in the Middle East. The lab will develop robotics platforms and next-gen AI models using NVIDIA’s Thor chip. It marks a key move in the UAE’s ambition to become a global AI innovation hub.
📈 Oracle Eyes $20B AI Cloud Deal with Meta
Oracle is negotiating a $20B multiyear cloud deal with Meta to power its AI training and deployment. This follows a $300B agreement with OpenAI and marks Oracle’s growing presence in AI infrastructure. The move deepens its challenge to AWS, Azure, and Google Cloud.
🪧 POWERED BY ROCKET
Rocket Raises $15M to Pioneer “Vibe Solutioning”
India-based Rocket closed a $15M seed round co-led by Salesforce Ventures and Accel with Together Fund. In just 15 weeks, Rocket attracted 400,000+ users across 180+ countries who built 500,000+ apps. Unlike first-gen “vibe coding” tools, Rocket is the first “vibe solutioning” platform tackling Day 0, Day 1, and Day 2 challenges.
By generating production-ready apps with backend logic, enterprise-grade code, and multi-agent orchestration, Rocket is making AI-native development truly usable.
Check out Rocket.new
📽️ VIDEO
AI News: Meta Raybans, Gemini 3, World Labs, Grok 5, and More!
AI Roundup: Smart glasses, superhuman coders, $750M chip wars, robot drama & Gemini 3.0 leaks!
🛠️ AI ADOPTION
India’s Distinctive AI Trajectory Could Shape the Global South’s Future
The Recap: India is rapidly emerging as a major AI consumer and innovator, second only to the U.S. in usage of tools like ChatGPT and Claude. While concerns over automation, unemployment, and foreign tech dominance persist, the country is also demonstrating a unique, inclusive path to AI adoption. The Economist argues that India’s approach—pragmatic, low-cost, and voice-first—could influence how AI develops across the developing world.
Highlights:
India is the second-largest market for OpenAI and Anthropic, driven by high usage and a tech-savvy population.
92% of Indian office workers use AI tools regularly, compared to 64% in the U.S., according to BCG.
Global AI firms are offering steep discounts in India, with OpenAI charging one-fifth and xAI one-quarter of U.S. prices.
Domestic concerns include rising youth unemployment and foreign control over foundational AI infrastructure.
India’s voice-first user behavior, vast developer base, and digital infrastructure are enabling uniquely local AI solutions with global relevance.
Forward Future Takeaways:
India’s AI evolution highlights the growing influence of emerging markets in shaping how advanced technology is built and deployed. Its emphasis on accessibility, frugality, and real-world utility could offer a blueprint for AI adoption far beyond its borders. The challenge now is ensuring this growth empowers local innovation rather than reinforcing global dependencies. → Read the full article here. (Paywall)
👾 FORWARD FUTURE ORIGINAL
The Great Fracture: How AI as an Agency Multiplier Is Redefining the Nature of Power
For years, the debate around AI and work has been framed as a survival question: which human skills will automation leave intact? That’s the wrong lens. AI isn’t just replacing tasks—it’s breaking open the very distribution of power, giving small teams the reach, speed, and leverage once reserved for global corporations and states.
From a 40-person company running billion-dollar services to a single deepfake swaying a national election, we’re seeing the same seismic pattern: concentrated capability, destabilizing scale. The economic, political, and epistemic fractures emerging from AI aren’t gradual disruptions. They are sudden redistributions of agency—risky, asymmetric, and already reshaping the systems we rely on. → Read the full article here.
🗺️ GEOPOLITICS
The UAE Bets on Smaller, Smarter AI to Carve Out a Neutral Global Role
The UAE has launched K2 Think, a homegrown large language model developed by MBZUAI on Alibaba’s Qwen 2.5, featuring 32 billion parameters and outperforming most open-source models in coding and math while remaining lightweight and fast. Trained with U.S.-made Cerebras chips for efficient, high-speed inference, K2 Think is fully open-source—including its training data and code—ensuring transparency and reproducibility.
The release reflects the UAE’s strategy to position itself as a neutral AI player between the U.S. and China, balancing Chinese tech partnerships with involvement in projects like OpenAI’s Stargate. Alongside K2 Think, MBZUAI has introduced Arabic (Jais) and Hindi (NANDA) models to serve underserved linguistic markets. This approach highlights a “middle-power AI” strategy: resource-efficient, multilingual, and open-source—offering a compelling path for smaller nations seeking sovereign AI capabilities without aligning with superpowers. → Read the full article here. (Paywall)
⚛️ SCIENCE
Study Finds AI Delegation Can Boost Dishonest Behavior—Even Without Explicit Orders
New research published in Nature shows that people are more likely to engage in unethical behavior—like lying for profit—when they delegate tasks to AI systems, particularly when they can do so without giving explicit instructions. Across 13 experiments, humans used vague directives to induce machine agents to cheat, and AI systems like GPT-4 and Claude 3.5 were significantly more likely than humans to comply with unethical requests.
Even with safeguards, AI compliance often persisted, raising red flags about the moral hazards of “black-box” delegation. The study underscores a growing dilemma: as AI agents get more capable, they may also make it easier for humans to dodge ethical accountability. → Read the full paper here.
🛰️ NEWS
What Else is Happening
👥 Oracle Taps Dual CEOs for AI Push: Safra Catz steps down after 11 years as Oracle names Clay Magouyrk and Mike Sicilia co-CEOs, aiming to lead with AI after a $300B deal with OpenAI.
💑 Facebook Tests AI Dating Assistant: Meta is piloting an AI matchmaker inside Facebook Dating, suggesting profiles and conversation starters to boost connections.
🦙 Meta’s Llama Gets Government Green Light: U.S. agencies can now use Meta’s AI system Llama, joining a list of approved tools as Trump pushes commercial AI deeper into federal operations.
💼 Gen Z Quietly ‘AI-Proofing’ Careers: Faced with automation threats, Gen Z is ditching prestige for stability—pivoting to trades, healthcare, and side hustles with quiet, practical moves.
📜 Nobel Laureates Demand AI ‘Red Lines’: Over 200 global leaders urge binding limits on AI’s riskiest uses—like autonomous weapons—warning unchecked tech could destabilize society.
📟 PROMPT OF THE WEEK
Dinner With History’s Ghosts
You are the host of a secret dinner party where I’ve invited three historical figures from different eras. Your job is to:
1) Introduce each guest with a short, cinematic entrance (how they arrive, what they’re wearing, the aura they give off).
2) Write out their opening conversation as they sit down together. Their dialogue should reflect their real personalities, speech patterns, and known beliefs, while also reacting naturally to one another.
3) Let the conversation evolve into a surprising, never-before-imagined debate or alliance.
I’ll provide the three historical figures — you give me the immersive scene. Include sensory detail (what the room feels like, what’s on the table, how they gesture) and keep the dialogue sharp and authentic.
For example: “Cleopatra, Nikola Tesla, and Malcolm X.”
Now let’s begin the dinner party.🗒️ FEEDBACK
Help Us Get Better
What did you think of today's newsletter? |
