AI News — Tuesday, September 22, 2026

10
OpenAI Forms Math Advisory Group as AI Resolves Over 100 Open Problems

OpenAI has established a new advisory group focused on mathematics, highlighting its AI's recent achievement of solving more than 100 previously open mathematical problems.

TechCrunchindustry
9
Higgsfield AI Ships New Video Features in a Day with GPT-6 Astra

Higgsfield AI rapidly developed and deployed new video features, showcasing the impressive speed and capability of OpenAI's GPT-6 Astra model in production environments.

OpenAI Blogproduct
8
IntBMoE: Integrating Block-Level Conditioning into Expert Composition for Full-Participation Mixture-of-Experts

A new research paper introduces IntBMoE, a method for integrating block-level conditioning into Mixture-of-Experts models to improve their efficiency and performance.

Hugging Faceresearch
8
Grounded Skill Synthesis from Code at Scale for Agentic Intelligence

Researchers propose a method for synthesizing grounded skills from code at scale, aiming to enhance the capabilities of agentic AI systems.

Hugging Faceresearch
8
Building Standards for the Next Phase of AI

OpenAI discusses the critical need for developing robust standards to guide the responsible and effective evolution of artificial intelligence.

OpenAI Blogindustry
7
What If Your AI Agent Never Had to Leave the Browser? (Demo 🚀)

A new demo explores the concept of AI agents operating entirely within a web browser, potentially simplifying deployment and enhancing user interaction.

Dev.toproduct
7
RecreationWorld: Scalable and Verifiable Environments for Hybrid Computer-Use Agents

RecreationWorld introduces new scalable and verifiable environments designed for training and evaluating hybrid AI agents that interact with computers.

Hugging Faceresearch
7
RRSI: Regularized Recursive Self-Improvement of Agent Harnesses

A new paper presents Regularized Recursive Self-Improvement (RRSI), a technique for enhancing AI agents through iterative self-improvement of their control mechanisms.

Hugging Faceresearch
6
The Man Who Built Apple's Stores Doesn't Buy Silicon Valley's Bet on AI Shopping

The former head of Apple's retail strategy expresses skepticism about Silicon Valley's current enthusiasm for AI-driven shopping experiences.

TechCrunchindustry
6
ChatGPT Now Knows What You Do on Other Websites via Ad Collector

Reports indicate that ChatGPT may be collecting user data from other websites through an ad collector, raising privacy concerns among users.

Lobste.rsindustry
6
OmniVChat: Synthesizing, Benchmarking, and Training for Native Audio-Visual Dialogue

OmniVChat introduces a new framework for synthesizing, benchmarking, and training models capable of native audio-visual dialogue, advancing multimodal AI.

Hugging Faceresearch
6
Paint-Anything: Unified Any-Color Control for Image Generation and Editing

A new research paper presents Paint-Anything, a unified approach that allows for flexible, any-color control in both image generation and editing tasks.

Hugging Faceresearch
6
Transferring the Intelligence of VLMs to Robotic Control

Researchers explore methods for effectively transferring the intelligence and capabilities of Vision-Language Models (VLMs) to enhance robotic control systems.

Hugging Faceresearch
5
How to Stop AI from Confidently Shipping Broken Code (A Pattern That Actually Works)

An article proposes a practical pattern to prevent AI from generating and confidently deploying faulty code, addressing a common challenge in AI-assisted development.

Dev.toproduct
5
A Continual Learning Model Trained from Scratch on 8GB VRAM Laptop with Batch-1 Stream of Data

A developer demonstrates a continual learning model capable of training from scratch on a modest 8GB VRAM laptop using a single-batch data stream, highlighting efficient AI development.

Lobste.rsopen-source