Monday, May 11, 2026

Start of archive·May 16

Top story

The NumbersThe Information

OpenAI Launches $10 Billion Private-Equity Joint Venture, Acquires Consultancy

OpenAI is shifting from pure model licensing to enterprise deployment services via a new JV structure—signaling a major revenue diversification play and potential path to profitability beyond API sales.

OpenAI Deployment Company: $10B private-equity joint venture announced

The briefs

Nvidia is shifting from pure hardware supplier to portfolio investor, directly funding the AI ecosystem that depends on its chips. This vertically integrates Nvidia's influence across model development, infrastructure, and applications—reshaping competitive dynamics.

$40B+ invested in AI companies in 2025

Chinese tech giants are weaponizing AI spinoffs to capture investor capital before going public. Kuaishou's move signals a new playbook: unbundle AI business units into standalone, high-valuation entities to maximize IPO readiness and hedge against regulatory risk.

Kuaishou spinning off Kling AI video unit

Cerebras's elevated IPO pricing signals strong investor conviction in specialized AI chip makers as an alternative to NVIDIA dominance. The $4.5B raise and upward revision reflect surging demand for custom silicon in the AI infrastructure race.

IPO pricing range raised from $115–$125 to $150–$160 per share

As GPU demand outpaces financing timelines, the AI buildout has shifted from a compute problem to a capital and execution choreography problem. Infrastructure providers are now racing to secure long-term customer contracts (especially hyperscaler deals) to unlock financing for multi-year data center builds.

Single gigawatt capacity estimated at $50B (Nvidia CEO Jensen Huang)

Cerebras's IPO signals investor appetite for AI-native chip companies positioned to challenge NVIDIA's dominance. The timing—with Elon Musk's historical interest in a merger—underscores how fragmented the compute layer is becoming as demand for specialized AI silicon explodes.

IPO range bumped up; targeting $4.8B raise

As NVIDIA scarcity forces AI labs into multi-chip strategies, a new service layer is emerging to handle chip-specific optimization. This reveals how the AI infrastructure stack is fragmenting beyond GPU monoculture—and creating new business opportunities in the plumbing.

OpenAI hired Gimlet Labs to optimize models for Cerebras chips

OpenAI restructured a revenue-sharing agreement that could have cost $135B, reducing exposure dramatically. This reshapes the economics of the AI leader's most critical partnership and signals shifting leverage in the AI industry's biggest power dynamic.

OpenAI saves $97 billion through 2030 via renegotiated Microsoft deal

Anthropic is shipping a structurally different go-to-market model on AWS where Anthropic retains platform operations while AWS handles billing/IAM. This signals a power shift in how AI vendors negotiate with hyperscalers and could reshape enterprise procurement patterns.

Claude Platform on AWS reaches general availability

A former OpenAI CTO's new startup is introducing a fundamentally different model architecture (interaction models) aimed at real-time, humanlike responses. This represents a potential shift in how AI systems handle conversational flow and could challenge the current turn-based paradigm dominated by OpenAI and Anthropic.

Mira Murati founded Thinking Machines Lab after leaving OpenAI CTO role