Issue #3
Vibe Coding Weekly Issue Three is out. Here's what you'll find in this issue:
- Lovable 2.0 launched with multiplayer workspaces, an AI coding assistant, security scanning, and easier app deployment features.
- Gian Segato explains why personal agency, not just skill or education, is becoming the most important factor for success in the AI era.
- Anthropic's CEO, Dario Amodei, warns that AI systems must be made interpretable before they become too complex to control safely.
And a few more picks to keep you sharp and inspired.
🛠️ Tools For Vibe Coders
SampleApp.ai is an AI-powered platform that helps developers and companies quickly generate customized, sales-ready demo apps using their own API or SDK documentation. With features like real-time sync, pre-built playgrounds, trend-driven app ideas, and custom training, it streamlines the creation of tailored sample apps for use cases like developer advocacy, sales demos, and onboarding. Designed to reduce development time, it reflects the modern shift in priorities—focusing on rapid sample app delivery over initial API calls.
📰 AI Industry News
Stay informed about the latest developments in the AI and development ecosystem.
Lovable 2.0 Released
Lovable introduces significant enhancements to its AI-powered platform, streamlining the process of building production-ready apps through conversational prompts. Key updates include:
- Multiplayer Workspaces: Collaborate with up to 20 users in shared workspaces, facilitating team-based development.
- Chat Mode Agent: An intelligent assistant that aids in planning, debugging, and navigating code without altering it.
- Security Scan: Integrated vulnerability detection for apps connected to Supabase, promoting safer development practices.
- Dev Mode & Visual Edits: Direct code editing and intuitive style adjustments within the platform.
- Custom Domains: Simplified process for connecting custom domains to applications.
These features aim to enhance usability, collaboration, and security, making Lovable 2.0 a more robust tool for developers and teams.
New Design Upgrades For Bolt.new Agent
Recent updates to Bolt.new agent have introduced significant design-focused enhancements, making the platform more powerful and user-friendly for web application development directly in the browser. The new design agent uses advanced reasoning to generate visually coherent, customized UI designs from prompts, thoughtfully considering color schemes, animations, and layout. It produces high-quality results, like timelines and scrollable sections, reducing the manual work typically needed in tools like Figma on the first try.
Google’s AI search numbers are growing, and that’s by design (TechCrunch)
Google's AI-driven search features, particularly AI Overviews, have grown significantly, now serving over 1.5 billion monthly users across more than 100 countries. These AI-generated summaries appear at the top of search results, providing concise answers to user queries. While some publishers express concerns about reduced web traffic, Google views these features as opportunities to enhance user engagement and introduce new advertising avenues. The company has also introduced "AI Mode," allowing users to engage in conversational searches, and expanded "Circle to Search," enabling users to highlight on-screen content and ask questions, now available on over 250 million devices. Additionally, visual searches via Google Lens have increased by 5 billion since October, with a notable rise in shopping-related queries.
📝 Featured Reads
Stay in the loop with standout articles from the world of AI, coding, and tech trends.
Agency Is Eating the World (Pirate Wires)
In this post, Gian Segato argues that the rise of AI tools has shifted the economic landscape, enabling individuals to launch billion-dollar businesses independently. He posits that the critical dividing line in today's economy is no longer skill or education but the will to act. This democratization of entrepreneurial power challenges traditional hierarchies and emphasizes personal agency in leveraging AI for business success.
The Urgency of Interpretability
In his essay, Anthropic CEO Dario Amodei stresses the importance of understanding how AI systems work internally before they become too advanced to control. He describes current AI models as black boxes whose decision-making processes are largely opaque, even to their creators. This lack of insight could lead to dangerous outcomes, such as unintended behaviors or manipulation. Amodei calls for deeper research into mechanistic interpretability, comparing it to building an MRI for AI that reveals how these systems think. He remains optimistic about recent progress but urges faster development to ensure AI remains safe and aligned with human values.
Vibe Coding is not an excuse for low-quality work
In his essay, Addy Osmani critically examines how people are increasingly turning to AI-generated code without really understanding it or keeping an eye on quality. He points out that while these AI tools can boost productivity, they shouldn't replace essential software engineering principles like making code easy to read, maintain, and thoroughly test. Osmani warns that getting too comfortable relying on AI can be a problem, and he suggests that developers should use AI help while still sticking to high standards to ensure they deliver solid and dependable software.
How Does ‘Vibe Coding’ Work With Observability?
In the Honeycomb blog post, Austin Parker explores integrating AI-assisted development, called "vibe coding," with observability practices. He emphasizes that while AI tools can accelerate coding by generating boilerplate and suggesting patterns, they often lack context about the specific application, leading to potential issues. To mitigate this, Parker advocates for embedding observability early in the development process, allowing developers to monitor and understand AI-generated code behavior in real time. By combining AI-driven coding with robust observability, teams can maintain code quality, quickly identify anomalies, and ensure that the software behaves as intended.
It’s all fun and games… until your boss starts vibe coding
In the article, Michael F. Buckley critiques the rise of Vibe Coding, where individuals use AI tools to rapidly generate digital products based on vague prompts, often without a deep understanding of design or development principles. He highlights the risks of such practices, including creating subpar products that may appear polished but lack functionality and scalability. Buckley warns that this trend can lead to stakeholders undervaluing the expertise of professionals and pushing for quick, low-quality solutions that ultimately harm the product's integrity. He calls for a return to thoughtful, well-crafted design and development processes to ensure the creation of meaningful and compelling digital experiences.
📚Guides & Tutorials
Building MCPs with Netlify
Netlify’s dev guide on writing Model Context Protocol (MCP) servers explains how to build and deploy AI-powered MCP endpoints using Netlify’s Agent Experience (AX) Functions. MCP enables developers to define how AI agents interact with external tools and APIs in a standardized, composable way. This setup helps developers create smarter, more context-aware AI experiences on the web with minimal infrastructure overhead.
📺 Videos Worth Watching
Tom offers practical advice for developers using AI tools in coding. He recommends beginner-friendly tools like Replet and Lovable for quick UI changes while suggesting advanced tools such as Windsurf and Cursor for more experienced users. He stresses the importance of version control, high-level integration tests, and using AI for non-coding tasks like DNS configuration and favicons. Tom advises thorough documentation and detailed instructions when working with AI, developing complex features in clean codebases, and leveraging GitHub references. He highlights the need for frequent refactoring, effective use of voice input and screenshots, and staying updated with new AI models for continuous improvement.
📅 Events
Upcoming hackathons or events.
Join top engineers, builders, and researchers at the biggest MCP hackathon! Build from scratch, tackle tough tech challenges, and push AI infrastructure forward.
$10K+ in prizes, rapid-fire sub-challenges, and one intense day to go from idea to demo—no pre-built projects allowed.
Join the Vibe Coding Hackathon by Sampleapp.ai for a hands-on, immersive experience where developers, founders, and tech enthusiasts can build innovative solutions using the latest AI tools and APIs. Collaborate with peers, get mentorship from industry leaders, and compete for exciting prizes in a creativity-fueled environment designed to inspire breakthrough ideas.
😂 Vibe Coding Jokes
We hope you like it. Until next time, keep vibing 👨💻!