Analytics DeepLearning.AI - The Batch | DeepLearning.AI (Page 12)

A robot wears a bow around its finger with SPEC.md attached to it. The developer asks "What's that for?"; The robot says "You told me something important last Tuesday. I'm never letting go."

Letters

Make All Your Tokens (and Your Brainwork) Count: Agentic loop coding works until it fulfills a goal. Here's an agentic way to iterate on your goals to find the right one.

The key idea of agentic coding loops is to have an agent keep working until it satisfies a condition, such as achieving a product specification.

The Batch Newsletter

Restoration of Claude Fable 5, Gemini's Video Dev Engine, DeepSeek Speeds Up Speculative Decoding

The Batch News & Insights: The key idea of agentic coding loops is to have an agent keep working until it satisfies a condition, such as achieving a product specification.

A perplexing airport customs scene where a robot is inspected by an officer, highlighting border security technology.

Data Points

GPT-5.6 will see a wider release this week: China’s government moves to control its companies’ AI products

Copilot will decrease reliance on GPT, Claude. Claude Cowork gets web, mobile implementations. Nvidia’s Audex blends audio and voice with text intelligence. OpenAI mini voice model promises to keep costs low.

Researcher digitizing newspapers on tablet for archival accuracy, in document storage room with organized boxes.

Data Points

Alibaba and ByteDance quash human-like bots: And, in case you missed it: some recent under-the-radar news

Seedance 2.5, the movie industry’s favorite video generator. Field-specific knowledge helps Claude code users most. DiffusionGemma, Google’s fast, capable text generator. Regular use of AI chatbots rises among Americans.

Robot arm successfully places pot on cloth, demonstrating reward verification and scoring effectiveness.

Machine Learning Research

Better Reward Models for Robots: Inside RoboReward, a family of vision-language reward models that train robots to take action

When you’re training a robot via reinforcement learning, a handcrafted reward function is labor-intensive to build but often dispenses rewards more effectively than a general-purpose reward model based on a vision-language model. Researchers built reward models that narrowed the gap.

The table shows MAI-Thinking-1 leading in several benchmarks, compared to other AI models.

Machine Learning Research

Microsoft Strikes Out on Its Own: Microsoft revealed MAI-Thinking-1, a Claude Sonnet 4.6-sized reasoning model developed without distillation

Microsoft, once OpenAI’s exclusive partner and still a major reseller of other companies’ AI models, built its own reasoning model from scratch.

Six charts show Fugu and Fugu Ultra scoring highest, marked by red bars, on various tasks and benchmarks.

Machine Learning Research

Fugu Blends Models Task by Task: Sakana debuted dedicated orchestrator models, Fugu and Fugu-Ultra, that spawn Claude, Gemini, and GPT agents

Models that orchestrate the activities of other models and agents achieved state-of-the-art performance on a variety of benchmarks, outperforming the best individual models working alone.

Machine Learning Research

GPT-5.6 Lands in Limbo: OpenAI previewed three GPT-5.6 Models (Sol, Terra, and Luna), wider release coming soon

OpenAI announced a preview of its GPT-5.6 family, including a top-tier model comparable to Claude 5 Mythos — but so far it’s available only to users that are selected by the U.S. government.

Cartoon features a man and woman contemplating three luxury sports cars: blue, red, and orange on a blue background.

Letters

How We Decide What Courses to Teach: The AI world is full of hype and sales pitches. DeepLearning.AI focuses on most important tools and techniques in ways you can apply to any AI vendor’s ecosystem.

The AI world has become incredibly noisy. Social media, traditional media, and an army of marketers produce a cacophony of hype and content that are often secretly sales pitches for their products.

The Batch Newsletter

OpenAI’s GPT-5.6 Family, New Ways to Train Robots, Models Invoking Models

The Batch News & Insights: The AI world has become incredibly noisy. Social media, traditional media, and an army of marketers produce a cacophony of hype and content that are often secretly sales pitches for their products.

Storyboard artist sketches movie scenes at desk, surrounded by script notes and team collaboration, for film planning.

Data Points

U.S. lifts export controls on Claude models: The mid-tier is back with Claude Sonnet 5

Claude Science, a workbench for biological research. Gemini’s Nano Banana 2 Lite and Omni Flash. Devin Fusion pairs a top model with an inexpensive sidekick. OpenClaw’s iOS and Android controller applications.

Machine Learning Research

Anthropic Opus 4.8 Leaps Forward: Claude Opus 4.8 won back the high-performance crown for Anthropic, pending wider availability of its Mythos-class models

Anthropic mid-2026 update of Opus held the the top of a leading intelligence ranking for about a week, only to be overtaken by Claude Fable 5.