Sign in Subscribe
2 min read

Open-Source LingBot-World Beats Genie 3 as OpenAI Retires GPT-4o, GPT-4.1

New AI Models and Frameworks

LingBot-World Outperforms Genie 3 in Dynamic Simulation and is Fully Open Source
The open-source LingBot-World framework has been released, surpassing Genie 3 in dynamic simulation performance with 16 FPS and emergent spatial memory. It breaks proprietary barriers by offering full access to code and model weights, enabling broader research and development in interactive world simulation.

AI Model Retirements and Updates

OpenAI Retires GPT-4o, GPT-4.1, and Older Models on February 13, 2026
OpenAI will deprecate GPT-4o, GPT-4.1, GPT-4.1 mini, and o4-mini in ChatGPT, shifting focus to newer models like GPT-5.2. The move aligns with OpenAI’s strategy to streamline offerings and prioritize more advanced AI systems.

Local AI and Open-Source Developments

OpenCode + llama.cpp + GLM-4.7 Flash Enables Local Claude Code Workflow
A user demonstrated running Claude Code-like functionality locally using OpenCode, llama.cpp, and GLM-4.7 Flash, with potential for further optimization via speculative decoding. This highlights the growing feasibility of advanced AI models in offline or private environments.

Mistral OCR Skill Enhances AI Agents for Document Conversion
A new OCR skill for AI agents like Claude Code and OpenCode leverages Mistral’s API to convert PDFs/images to Markdown, improving speed and accuracy over default methods. This tool addresses common pain points in document processing workflows.

AI Ethics and Governance

Pentagon and Anthropic Clash Over Military AI Safeguards
Negotiations between the Pentagon and Anthropic have stalled over AI usage restrictions, with the military seeking to remove safeguards for autonomous weapons targeting and domestic surveillance. Anthropic insists on maintaining ethical limits on its technology’s deployment.

Industry Perspectives and Vision

Mistral CEO Advocates for AI as a Utility Like Electricity
Mistral CEO Arthur Mensch compared intelligence to electricity, emphasizing the need for unrestricted access to AI to prevent throttling by gatekeepers. This aligns with Mistral’s push for open and accessible AI models.