Weekly AI news for engineers, executives, and enthusiasts.

Nvidia’s enterprise-focused NemoClaw gives OpenClaw a security boost: Claude Dispatch lets paid users authorize remote work

GPT-5.4 Mini and Nano, optimized for speed. Nvidia’s own tiny-sized open-weights model. Midjourney updates to v8 in Alpha. Mamba releases third version of its state space language model.

Perplexity Computer hits desktop, mobile, Pro: Google releases Aletheia, its top natural language math agent

DeepSeek testing favors local chipmakers. Nvidia’s reasoning-added retrieval method. Why OpenAI and Oracle shut down Stargate expansion. Meta’s giant deal with AI cloud provider Nebius.

Lightning-Fast Diffusion Learning: Inside Feature Auto-Encoder, a diffusion image generator that shrinks embeddings for more speed

Research shows that diffusion image generators can train somewhat faster if they learn to reconstruct embeddings from a pretrained encoder that’s built for vision tasks like classification, segmentation, and retrieval — not image generation.

AI Data Centers Go Off the Grid: Rather than rely on public utilities, AI companies build their own power plants

Meta and OpenAI are among the tech companies that are building private power plants that will operate independently of regional grids to supply electricity for their massive buildout of AI data centers.

AI on Mobile Skyrockets: State of Mobile 2026 Report shows AI chatbot, search, and assistant growth outpaces gaming, social, and more

Downloads of mobile AI apps and resulting revenue are surging.

GPT-5.4’s Higher Performance, Higher Price: OpenAI’s GPT-5.4 Pro and GPT-5.4 Thinking challenge Google’s Gemini 3.1 Pro Preview as best all-around AI model

OpenAI updated its flagship models, extending the ability to use tools and setting the state of the art on a handful of benchmarks, and priced them at the top of the market. Its coding and agentic abilities have enabled Codex, OpenAI’s competitor to Anthropic’s Claude Code, to leap ahead.

Let’s Help Agents Share Their Work: AI Social Networks Can Be More than Fun and Games

Should there be a Stack Overflow for AI coding agents to share their learnings with each other?

GPT-5.4 Makes A Splash, AI’s Growth on Mobile, Data Centers Go Off-Grid, Apple’s Diffusion Research

The Batch AI News and Insights: Should there be a Stack Overflow for AI coding agents to share their learnings with each other?

DeerFlow 2.0 puts new spin on Claw-like agents: Anthropic sues the U.S., which shows no signs of backing down

Meta’s purchase of Moltbook, a social network for AI agents. Iran’s attacks on data centers in the UAE and elsewhere. Google’s new multimodal embedding model. Grammarly’s AI ventriloquism using famous writers.

GPT-5.4 Pro challenges Gemini 3.1 Pro Preview: Luma’s new state-of-the-art image editor/interpreter

Microsoft’s new open-weights vision reasoning model. Yuan 3.0 Ultra, a document-retrieving juggernaut. OpenAI’s hardware leader’s resignation. Black Forest’s new training method for image models.

Context for Coding Agents: Agentic coding systems often make mistakes because they’re not aware of tools, API calls, and the like that came out after they were trained. Context Hub gives them the documentation they need to write correct code.

I’m thrilled to announce Context Hub, a new tool to give to your coding agents the API documentation they need to write correct code.

Agent Solves Stubborn Math Problems: Google’s Aletheia uses Gemini 3 Deep Think to find original mathematics solutions

LLMs have achieved gold-medal performance in math competitions. An agentic system showed strength in mathematical research as well.