Machine Learning Research
Your Robot Dev Team: OpenAI introduces Codex, a multi-agent cloud-based software engineering tool in ChatGPT
OpenAI launched an agentic software-development system.
Latest
Data Points
Radiologists use AI to automate tasks, not jobs: Economists say data for landmark paper was flawed
U.S., China spar over Huawei chips. Nvidia makes it easier to build custom data centers. Meta’s Open Molecules project hopes to revolutionize chemistry. New research examines language models’ win rates in games.
Data Points
DeepSeek outlines V3 training, hardware limits: OpenAI’s Codex now assists with code in the cloud
Windsurf introduces SWE-1 family of coding/engineering models. Stripe adapts transformer architecture for versatile payments model. Alibaba’s top video model gets another boost. U.S. Republicans make an end run around local AI regulations.
Letters
The Value of AI’s Speed Is Underrated: The value of AI-driven productivity is well recognized, but AI-driven speed can create growth, too.
AI’s ability to make tasks not just cheaper, but also faster, is underrated in its importance in creating business value.
The Batch Newsletter
Recipes For Reasoning, Open and Compact Code Generator, Looser AI Regulations, More Factual Output
The Batch AI News and Insights: AI’s ability to make tasks not just cheaper, but also faster, is underrated in its importance in creating business value.
Machine Learning Research
Memory Layers for More-Factual Output: Meta researchers build Llama-style models that recall details without needing more computing resources
Improving a large language model’s factual accuracy typically requires making it bigger, which in turn, involves more computation. Researchers devised an architecture that enables models to recall relevant details without significantly increasing the amount of computation required.
Business
EU Loosens AI Regulations: European regulators move to relax some AI Act rules on developers’ liability, other provisions
The European Union made an abrupt U-turn away from its stringent AI regulations. Meta promptly adjusted to the loosening restrictions.
Machine Learning Research
Open, Compact Code Generator: DeepCoder-14B-Preview further fine-tunes reasoning models for coding
An open-source code generator performs comparably to the reasoning models DeepSeek-R1 and OpenAI o1 with a much smaller model.
Machine Learning Research
Reasoning Models With Recipes: Microsoft unveils training details for Phi-4-reasoning, Phi-4-reasoning-plus, and Phi-4-mini-reasoning
Microsoft published its latest recipe for training reasoning models, substantially expanding what is still a fairly small base of public knowledge.
Data Points
Report says common AI model training practices may violate current U.S. copyright law: OpenAI’s CLIP topped by new open-source vision encoders
Microsoft joins Google’s Agent2Agent project. Police and governments circumvent face-tracking laws. OpenAI and Microsoft reportedly seek a new deal. Researchers use RL to train coding models starting with zero data.
Data Points
Updated Gemini Pro model builds interactive websites from prompts: OpenAI unveils new restructuring plan
Mistral’s new medium-sized language model. Claude developers gain access to web search. Alibaba uses RL to teach LLMs to search better. OpenAI’s plans to build AI infrastructure worldwide.
Letters
Hot Tips for Speedy Startups: Speed is the most important factor in successful startups. Here are four ways to accelerate your company.
I’m delighted to announce that AI Fund has closed $190M for our new fund, in an oversubscribed round.