Luis Poveda's AI Newsletter: Jun 30, 2025

In partnership with

Stop Asking AI Questions, and Start Building Personal AI Software.

Feeling overwhelmed by AI options or stuck on basic prompts? The AI Fast Track is your 5-day roadmap to solving problems faster with next-level artificial intelligence.

This free email course cuts through the noise with practical knowledge and real-world examples delivered daily. You'll go from learning essential foundations to writing effective prompts, building powerful Artifacts, creating a personal AI assistant, and developing working software—all without coding.

Join thousands who've transformed their workflows and future-proofed their AI skills in just one week.

Get the course

This week marked a significant shift toward more actionable AI tools, with ElevenLabs launching its voice-first 11ai personal assistant that can execute tasks through natural conversation, Google releasing the open-source Gemini CLI to bring AI directly into developers' terminals, Anthropic transforming Claude Artifacts into an AI app creation platform reminiscent of Flash gaming portals, and Chinese company MiniMax unveiling its M1 reasoning model that challenges Western AI dominance with competitive performance and cost efficiency.

ElevenLabs Launches 11ai: The Voice Assistant That Actually Takes Action

ElevenLabs has introduced 11ai, a groundbreaking voice-first conversational AI assistant that moves beyond traditional voice assistants by taking meaningful actions within user workflows. Powered by ElevenLabs Conversational AI and integrated with the Model Context Protocol (MCP), 11ai connects to various tools and systems to perform tasks such as planning, research, project management, and communication. The assistant delivers context-aware assistance through Retrieval-Augmented Generation (RAG), ensuring responses align with user intent while enabling hands-free productivity through natural voice interaction. Currently in alpha release, 11ai enters a competitive field alongside voice assistants from Perplexity, Amazon, and Google, positioning itself as a productivity-focused solution that emphasizes actionable outcomes over simple conversation.

• Geekflare: This AI Assistant from ElevenLabs Can Plan Your Day with Voice Commands

• The Decoder: ElevenLabs launches 11ai, a voice assistant that uses MCP to integrate with digital workflow tools

Google Opens the Terminal Gates with Free Gemini CLI for Developers

Google has launched Gemini CLI, a fully open-source AI agent that brings the power of Gemini 2.5 Pro directly into developers' terminals. The command-line tool offers industry-leading usage allowances at 60 model requests per minute and 1,000 requests per day at no charge through a free Gemini Code Assist license. Built on emerging standards like the Model Context Protocol and featuring built-in Google Search integration, Gemini CLI enables developers to write code, debug issues, and streamline workflows through natural language commands. The tool represents Google's commitment to making AI more accessible to individual developers while maintaining the terminal's efficiency and ubiquity that developers rely on. Integration with Gemini Code Assist extends these capabilities into VS Code, creating a unified AI-powered development experience across both command-line and IDE environments.

• Google Blog: Gemini CLI: your open-source AI agent

• 9to5Google: Gemini CLI brings Google's AI agent to your terminal for free

Claude Artifacts Transforms Into an AI App Gallery with Flash Game Vibes

Anthropic has expanded its Claude Artifacts system into a personal AI app gallery that enables users to create interactive web applications through plain English dialogue. The updated platform allows artifacts to interact with Claude itself through a new "window.claude.complete()" function, enabling in-app conversational features like dynamic NPCs or tutors. Users can build and share browser-based applications that run on modern JavaScript and React code, creating self-contained experiences reminiscent of the Flash gaming era but with contemporary web technologies. The platform organizes example apps into categories like "Learn something," "Life hacks," and "Be creative," featuring tools ranging from interactive writing editors to 3D office simulators. Unlike traditional web development, these applications operate in a sandboxed environment that can only communicate with Claude itself, creating a simplified development experience where users describe ideas and Claude handles both interface code and AI logic.

• Ars Technica: Anthropic summons the spirit of Flash games for the AI age

• Inc: Anthropic Just Made Vibe Coding With Claude Absurdly Easy

Chinese AI Company MiniMax Challenges Western Models with Open-Source M1

Shanghai-based MiniMax has launched M1, positioning it as the world's first open-source, large-scale, hybrid-attention reasoning model that claims performance on par with leading international models while offering superior cost efficiency. The company asserts that M1 outperforms all Chinese closed-source models and demonstrates particular advantages in complex, productivity-oriented scenarios including software engineering, long context processing, and tool integration. When performing deep reasoning with 80,000 tokens, M1 reportedly requires only about 30% of the computing power needed by competing models like DeepSeek R1. The release represents part of a broader product launch strategy from MiniMax, which also introduced the video generation model Hailuo 02 and the general intelligent agent MiniMax Agent within the same week, signaling the company's ambition to compete across multiple AI application areas in the increasingly competitive global landscape.

• China Daily: Shanghai AI company launches open-source model

• CometAPI: What is MiniMax-M1? All You Need to Know

Beyond Silicon Valley: The Global AI Surge

We are undoubtedly living in extraordinary times, the era of AI. Every week brings a wave of news: new product launches, powerful models, and groundbreaking tools. It’s simply astonishing. Many agree that this is one of the best moments in history to start a company. But beyond entrepreneurship, it’s also a time to unlock real, transformative value from technology.

And while much of the spotlight shines on the U.S., innovation is thriving globally. Europe is producing standout companies like ElevenLabs, Lovable, Hugging Face, and Mistral AI. Meanwhile, China is rapidly closing the gap in the AI race, achieving remarkable progress with fewer resources, less capital, and restricted access to key hardware. DeepSeek is just one early example of what's coming.

This week's developments showcase a maturing AI landscape where tools are becoming more actionable and accessible across different use cases. From voice-controlled productivity assistants to open-source developer tools and creative platforms, the focus has shifted from impressive demos to practical applications that integrate seamlessly into existing workflows. The emergence of competitive alternatives from Chinese companies like MiniMax also signals that AI innovation is becoming truly global, with different regions developing distinct approaches to model architecture and deployment strategies. As these tools continue to evolve, users can expect more specialized, efficient AI solutions that prioritize real-world utility over raw capabilities.

Start learning AI in 2025

Keeping up with AI is hard – we get it!

That’s why over 1M professionals read Superhuman AI to stay ahead.

• Get daily AI news, tools, and tutorials

• Learn new AI skills you can use at work in 3 mins a day

• Become 10X more productive

Sign up and start mastering AI