Key launches, research breakthroughs, and policy moves shaping AI development and deployment this week.
Startup Recursive released Flint, a free, laptop-sized coding model that was primarily created using its own AI research system.
Shows how AI can automate model development, a skill useful for building custom coding assistants.
Source: TechCrunch AI
NVIDIA demonstrated that its Kumo Relational system can achieve higher accuracy than traditional supervised models on relational datasets without any training.
Highlights the potential of zero-training models for rapid deployment on structured data tasks.
Source: Hugging Face
Anthropic released Claude Duet as a research preview, positioning it as an AI agent aimed at small businesses and professional workflows.
Provides a concrete example of next-generation AI agents that can be integrated into productivity tools.
Source: The Verge AI
Walmart announced a program that enables AI agents to make purchases on behalf of users, with controls for spending limits and purchase approval.
Illustrates real-world deployment of AI agents in e-commerce, raising design and safety considerations.
Source: The Verge AI
OpenAI partnered with Google Cloud to offer its frontier models and Codex workflows through the Model Garden on the Cloud Marketplace.
Enables developers to access powerful models directly within cloud infrastructure, simplifying integration and scaling.
Source: OpenAI
Study how to build, fine-tune, and deploy AI agents using large language model APIs and cloud services. Understanding model integration, safety controls, and zero-training techniques will let you create agents like Claude Duet or Walmart's shopping bots.
Build my AI Engineer plan