Weekly AI news for engineers, executives, and enthusiasts.

OpenAI’s Answer to Gemini 3, Runway’s Interactive Worlds, Disney’s Alliance With OpenAI, Adapting LLMs for Low-Data Domains

The Batch AI News and Insights: As amazing as LLMs are, improving their knowledge today involves a more piecemeal process than is widely appreciated.

Gemini makes Deep Research available via API: Disney and OpenAI form a Sora licensing deal

Nvidia’s Nemotron 3 family of models, including a 30B standout. Microsoft’s RL-powered Agent Lightning. Google’s new Flash voice models and translation agents. “Slop,” 2025’s word of the year.

Anthropic, OpenAI donate open-source projects: GPT-5.2 arrives with new document creation abilities

Mistral’s Devstral 2’s cost-effective coding. GLM-4.6V’s addition of native function calling to vision-language reasoning. Cursor 2.2’s debug and web developer modes. A new executive order preempting local AI regulations in the U.S.

Build an Autonomous Agent Using This Simple Recipe!: The ability of large language models to carry out multiple steps autonomously makes it possible to build a capable agent in a few lines of code.

If you have not yet built an agentic workflow, I encourage you to try doing so, using the simple recipe I’ll share here!

Claude Opus 4.5 Saves Tokens, White House Boosts AI-Powered Science, Amazon Exposes Nova 2 Pro Checkpoints, Small Models Solve Hard Puzzles

The Batch AI News and Insights: If you have not yet built an agentic workflow, I encourage you to try doing so, using the simple recipe I’ll share here!

Small Models Solve Hard Puzzles: Tiny Recursive Model beats larger competitors at games like Sudoku and Maze

Large language models often fail at puzzles like Sudoku, for which a solution includes multiple elements and a single mistake invalidates all of them. Researchers showed that a tiny network, by repeatedly refining its solution, can solve this sort of puzzle well.

Amazon Steps Forward: Nova 2 family boosts cost-effective performance, adds new agentic features

Amazon raised the competitive profile of its foundation models and added services for custom model training and an agent platform for browser automation.

White House Orders AI for Science: Genesis Mission would share U.S. data and resources with top AI companies

President Trump launched a United States effort to use AI to speed up scientific breakthroughs.

Claude Does More With Fewer Tokens: Claude Opus 4.5 retakes the coding crown at one-third the price of its predecessor

Claude Opus 4.5, the latest version of Anthropic’s flagship model, extends the earlier version’s strengths in coding, computer use, and agentic workflows while generating fewer tokens.

Gemini 3 adds compute-intensive Deep Think: Two famous math problems solved by 10-person AI startup

An executive order that would challenge state AI regulations in the U.S. An 8B parameter open model that shines at coding and math tasks. Publishers’ copyright and trademark lawsuits against Perplexity. Gemini 3 Pro’s ability to visually recognize handwriting from any period.

Mistral 3 update adds four new open models: Amazon Nova 2 family competes with low prices, new agents

Nvidia’s open VLA reasoning model for self-driving cars. Hugging Face’s Claude skills pack that fine-tunes language models. LangChain’s LangSmith no-code, all-chat agent builder. MCP Blockly, a project for students to build their own MCP servers.

Why People Don’t Trust AI and What To Do About It: Recent surveys by Edelman and Pew Research show that Americans distrust AI. The AI community should take this seriously and work to regain public trust.

Separate reports by the publicity firm Edelman and Pew Research show that Americans, and more broadly large parts of Europe and the western world, do not trust AI and are not excited about it.