Hello from The AI Night,

Today in AI:

Here's the deal: Anthropic made the Claude Platform on AWS generally available, giving AWS customers full access to native Claude API features with AWS authentication and billing. New features ship the same day they go live on the native API.

The Breakdown:

The bigger picture: Anthropic built a front door inside AWS's house. Same IAM login, same CloudTrail logs, same invoice, but Anthropic controls the infrastructure and the customer relationship. Enterprise teams get full API parity without leaving AWS. Amazon gets the compute bill. Anthropic gets everything else. That is a partnership where only one side is building leverage.

Here's the deal: OpenAI launched the OpenAI Deployment Company, a standalone unit embedding AI engineers directly into enterprise organizations. It's also acquiring Tomoro, an applied AI firm, bringing roughly 150 Forward Deployed Engineers from day one.

The Breakdown:

The bigger picture: OpenAI spent five years saying the model is the product. Now it is hiring 150 engineers to sit inside client offices because the model alone was not enough. Four billion dollars in day-one capital means this is not an experiment. OpenAI just admitted the real competition was never Anthropic or Google. It was Accenture.

Here's the deal: NVIDIA launched a space computing platform to run AI directly in orbit. The company is partnering with AetherFlux, Axiom Space, Kepler, Planet Labs, Sophia Space, and Starcloud to deploy accelerated computing across orbital and ground environments.

The Breakdown:

The bigger picture: Satellites today beam raw data to Earth, process it on the ground, then send results back. That round trip is the bottleneck for every commercial imagery company. NVIDIA is doing what it always does. Arriving at a new compute layer before anyone else realizes it needs one. Six launch partners means the land grab is already over.

In our new white paper, The Architecture Behind AI-Native Revenue Automation, Tabs CTO Deepak Bapat breaks down what it actually takes to apply AI to revenue workflows without breaking the books.

You’ll learn why probabilistic reasoning isn’t enough for finance, how Tabs pairs LLMs with deterministic logic, and why a unified Commercial Graph is the foundation for scalable, audit-ready automation. From contract interpretation to cash application, this paper goes deep on where AI belongs—and where it absolutely doesn’t.

If you’re evaluating AI for billing, collections, or revenue operations, this is the architecture perspective most vendors won’t show you.

Get the whitepaper

Thinking Machines Labs released a research preview of "interaction models," a 276B parameter MoE architecture trained from scratch for real-time multimodal collaboration, achieving state of the art combined scores in intelligence and responsiveness benchmarks.

Cursor launched a Microsoft Teams integration that lets users mention @Cursor in any channel to delegate coding tasks to a cloud agent and generate pull requests automatically.

Anthropic released Claude's Constitution as an audiobook narrated by co-authors Amanda Askell and Joe Carlsmith, featuring a Q&A on its writing process and how the document may evolve with future models.

Lovable launched an aesthetics update that lets users set typography, layout, and color preferences, preview design concepts before building, and generate more polished landing pages, apps, and blogs.

OpenAI launched an OpenAI Developers plugin for Codex that creates project API keys, diagnoses API errors, and provides setup guidance for building AI apps and agents inside the coding agent.

That’s it for today’s edition of The AI Night.

Our goal is to cut through the noise, surface what actually changed, and explain why it matters.

2 ways to support us: