Luis Poveda's AI Newsletter: April 14, 2025

From agent communication protocols to memory upgrades - this week AI breaks down the walls of isolation

Author

Luis Poveda
April 14, 2025

Executive Summary

This week saw major developments in AI collaboration and interoperability, with Google unveiling its Agent to Agent (A2A) protocol and GitHub releasing its official MCP server. Meanwhile, the business world continues to embrace AI in new ways, with Shopify's CEO making waves with a bold AI-first hiring policy. User experience improvements weren't far behind as ChatGPT received a significant memory upgrade.

AI Development Tools

Google Unveils Firebase Studio: An AI-Powered Development Environment for Full-Stack Apps

Google has officially launched Firebase Studio, a cloud-based AI-powered integrated development environment that helps developers build and ship production-quality full-stack AI apps. The new platform unifies Project IDX with specialized AI agents and assistance from Gemini to provide a collaborative workspace accessible from anywhere. Firebase Studio supports rapid prototyping with natural language, offering both coding environments and a unique "App Prototyping agent" that can generate entire apps from multimodal prompts including text, images, and drawings.

"Firebase Studio represents our commitment to making app development more accessible and efficient for developers of all experience levels. We've reimagined the entire developer experience to eliminate friction points and accelerate the journey from idea to production".

Google Cloud Executive

AI Interoperability & Standards

GitHub Launches Official MCP Server for AI Model Integration

GitHub has released its official Model Context Protocol (MCP) server in public preview, taking ownership of what was previously an Anthropic-created protocol. The new open-source server has been rewritten in Go, improving its usability while maintaining 100% of the original server's functionality. New features include the ability to customize tool descriptions, support for code scanning, and a new get_me function for improved natural language user experience. MCP enables AI models to call functions, look up data, and interact with external systems through a standardized interface.

"MCP tools give LLMs a standardized way to call functions, look up data, and interact with the world. Anthropic created the protocol and built the first GitHub MCP server, which grew to be one of the most popular MCP servers in the expanding ecosystem. We are excited to take ownership of the server and continue its development".

GitHub Blog

Google Unveils Agent2Agent Protocol for AI Interoperability

Google has launched Agent2Agent (A2A), an open protocol designed to enable AI agents to communicate with each other across different frameworks and vendors. Supported by over 50 technology partners including Atlassian, Box, Salesforce, and SAP, the protocol facilitates secure information exchange and coordinated actions across enterprise platforms. A2A complements Anthropic's Model Context Protocol (MCP) and follows five key design principles: embracing agentic capabilities, building on existing standards, prioritizing security, supporting long-running tasks, and remaining modality-agnostic for various types of communication.

"The A2A protocol establishes a common language that will allow AI systems to communicate, collaborate, and solve problems together in ways that weren't possible before. This isn't just about making existing systems talk to each other—it's about enabling entirely new categories of multi-agent applications".

Google AI Research Director

AI Business & Policy

Shopify CEO Mandates AI Usage Before New Hiring Approvals

In a bold policy shift, Shopify CEO Tobi Lütke has instructed teams to prove that AI cannot perform a job before requesting additional headcount. In a company memo shared publicly on social media, Lütke wrote: "Before asking for more headcount and resources, teams must demonstrate why they cannot get what they want done using AI." The policy reflects Shopify's aggressive embrace of AI as a core business strategy, with the company adding questions about AI usage to performance reviews and emphasizing that AI competency will be a formal part of hiring decisions moving forward.

"AI isn't just another tool for us—it's the foundation of Shopify's next chapter in empowering entrepreneurs. We're investing heavily in AI because we see it as the greatest lever for democratizing commerce since the internet itself".

Tobi Lütke, Shopify CEO

Andrew Ng Warns US Tariffs Could Harm AI Innovation

Speaking at Davos 2025, artificial intelligence pioneer Andrew Ng expressed concerns over US export controls and recent tariffs, warning they could be damaging to both the US and global AI development. Ng specifically highlighted concerns about placing countries like India in Tier 2 for AI chips and technology access prioritization. The impact of these trade policies has rippled through the tech sector, with major companies seeing stock value declines and industry leaders expressing worries about potential AI innovation slowdowns.

"These tariffs risk creating artificial barriers to innovation at a time when global collaboration in AI is more important than ever. The complex supply chain for AI hardware means that tariffs don't just impact costs—they can fundamentally alter where and how AI research progresses".

Andrew Ng

AI User Experience

OpenAI Expands ChatGPT's Memory to Reference All Past Conversations

OpenAI has rolled out a significant update to ChatGPT's memory capabilities, enabling the AI assistant to reference information from all past conversations with a user. The new feature, called "reference saved memories," allows ChatGPT to tailor its responses based on previous interactions, making conversations more context-aware and personalized. The update will initially be available to ChatGPT Pro and Plus subscribers, except for those in regions requiring additional regulatory compliance like the EU and UK. Users maintain control with options to disable the feature, manage specific saved memories, or use temporary chats for conversations they don't want stored.

"The improved memory capabilities represent our ongoing commitment to making AI assistants that truly understand and adapt to each user's unique context. What we've built is more than just conversation history—it's a system that learns how to prioritize and utilize information about you and your preferences over time".

OpenAI Product Lead

Conclusion

This week's developments highlight a clear industry trend toward greater AI integration and interoperability. The introduction of Google's A2A protocol and GitHub's MCP server point to a future where AI systems can communicate and collaborate seamlessly. Meanwhile, business leaders like Shopify's CEO are pushing organizations to fundamentally rethink workflows with AI at the center. As these systems become more interconnected and gain improved memory capabilities like ChatGPT's latest update, we're witnessing the emergence of a more cohesive, contextual AI ecosystem that can better understand and serve users over time.

The Author

Luis Poveda

Luis Poveda’s AI Newsletter

Luis Poveda is a technology optimist and passionate innovator, constantly exploring and researching the latest trends. Based in Barcelona, he is currently focused on AI and developing a modern AI-driven network observability tool. He is also the creator and maintainer of Luis Poveda's AI Newsletter, where he curates and shares key insights on the evolving AI landscape.