It’s Tuesday
Here’s what’s worth your time today.
Nvidia’s chip performance is improving so quickly that it’s starting to run into a more fundamental constraint: physics. CEO Jensen Huang says the company expects to generate at least $1 trillion from its latest chips by 2027, driven by massive demand from data centers powering AI.
The challenge is energy. Every gain in AI capability requires more electricity and cooling, turning efficiency into the key bottleneck. Even as chips become more powerful, total energy usage keeps rising — meaning the faster AI scales, the more infrastructure it demands.
Supported by Lightfield
For most founders, that’s 5-7 hours a week. Not selling. Not building. Typing notes into fields so a database stays current.
Lightfield eliminates the trade-off. Every email, call, and meeting captured automatically. Follow-ups surface the moment a call ends. And when you need work done — “draft a check-in for every prospect who’s gone quiet” — it does it, pulling context from real conversations.
4-minute setup. Connect your email and calendar.
Use code CTRLALTDEL23 for 3 months off
OpenAI has reportedly partnered with Amazon Web Services to sell its AI products to U.S. government agencies, including for classified work. The move builds on its recent Pentagon deal and significantly expands its presence across federal systems.
The partnership puts OpenAI directly into territory dominated by Anthropic, whose models are deeply integrated into AWS. By leveraging AWS’s government infrastructure, OpenAI gains faster access to public-sector contracts — often seen as a signal of trust for broader enterprise adoption.
Cathie Wood of ARK Investment Management says AI is already driving productivity gains, pointing to 2.8% growth today and predicting it could reach up to 6% annually as adoption accelerates. She highlighted rapid revenue growth across AI leaders like OpenAI and Anthropic as early signs of that shift.
Wood believes AI could generate $10–12 trillion in new revenue over the next decade, positioning it as a major driver of global economic growth despite skepticism around short-term impact.
Nebius Group plans to raise $3.75 billion in debt to fund data center expansion and AI chip purchases, following its massive infrastructure deal with Meta. The company will issue convertible notes due in 2031 and 2033 as it scales capacity for AI workloads.
The move reflects a broader trend, with AI infrastructure players increasingly turning to debt to finance growth. But it’s also raising concerns about a potential bubble, as companies like Nvidia continue pouring billions into the fast-growing “neocloud” space.
World, co-founded by Sam Altman, has introduced a new tool called AgentKit to verify that AI shopping agents are acting on behalf of real humans. The system uses World ID, which is generated through biometric verification, to confirm a user is behind an agent’s actions.
The launch comes as “agentic commerce” grows, with companies like Amazon and Mastercard enabling AI-driven purchases. By linking identity to transactions, World is positioning itself as infrastructure for trust in a future where AI agents increasingly browse and buy on behalf of users.