🧠 Google's Agent2Agent protocol 🧠
Making AI agents truly collaborative

Google has unveiled Agent2Agent (A2A), an open protocol for AI agent interoperability developed in collaboration with over 50 companies including Atlassian, MongoDB, PayPal, Salesforce, and SAP. This breakthrough allows AI agents built with different technologies or by different vendors to securely communicate, coordinate, and exchange information. A2A supports quick tasks or long-duration processes, including multimodal interactions (text, audio, video), with agents exposing their capabilities via "Agent Cards" to collaborate on shared tasks. This represents a crucial step toward making AI agents function as true digital colleagues capable of real-time collaboration.

🤖 Make AI Agents launch 🤖
Automation without coding

Make has introduced Make AI Agents, intelligent automation assistants that process information and make decisions in real-time within the Make platform. These agents can be deployed across workflows and programmed using natural language prompts without coding requirements. Make AI Agents connect to over 2,000 integrations and 30,000 actions, enabling use cases like automated ticket routing and invoice verification. The feature is currently available in beta to select organizations as development continues, allowing early adopters to experiment with creating customized AI assistants for various business processes.

🛠️ PySpur 🛠️
Open-source AI Agent workflows

PySpur is a new open-source library that enables creating AI Agent workflows through a drag & drop interface. It simplifies the creation, testing, and deployment of agents, reducing development time. PySpur installs in seconds with pip, allows adding custom tools, and exports agents in JSON format, making it accessible for developers of all skill levels to build sophisticated AI agent solutions.

INTERESTING NEWS

🔍 Meta to train AI on European users' public content 🔍

Meta has announced plans to begin training its AI models on public content shared by adults on its platforms in the EU, as well as on user interactions with Meta AI. Starting this week, EU users will receive notifications about this data usage and can easily opt out through an objection form. Meta emphasizes that private messages won't be used for training, and data from users under 18 will be excluded. The company states this approach will help their AI better understand European cultures, languages, and regional nuances, following similar practices already implemented by competitors like Google and OpenAI.

🎬 Netflix tests AI-powered content search 🎬

Netflix is testing an OpenAI-powered search feature that recommends content based on users' moods. Currently available to iOS users in Australia and New Zealand, the company plans to roll out the feature to US users in the coming weeks. This innovation aims to enhance content discovery and personalization, making it easier for subscribers to find shows and movies that match their current preferences.