Tags: vibe coding

A non‑programmer used Claude, a large‑language model, to create a web app that logs personal encounters with administrative hassles. By stitching together Supabase, GitHub and Netlify, the maker turned a vague idea into a functional dashboard that rewards users with quotes and images. The project exposed security gaps, such as an exposed API key, and highlighted how AI can lower the barrier to software creation while still demanding basic technical chores. The app now serves a small community that shares grievances about everything from phone‑tree loops to subscription traps. Lire la suite

Lovable has released a mobile version of its AI-driven vibe coding platform on iOS and Android, letting users create web apps and sites by typing or speaking prompts. The app syncs projects with the desktop service, supports asynchronous builds, and lets developers inspect and tweak generated code. By keeping outputs web‑based, Lovable sidesteps recent App Store restrictions that have forced similar tools to redesign their delivery models. The launch aims to extend Lovable’s on‑the‑go workflow while preserving the core promise of turning natural language into functional software. Lire la suite

Apple’s App Store has seen an unprecedented influx of new apps created with AI‑driven “vibe coding” tools, driving an 84% jump in submissions in a single quarter. The surge has stretched Apple’s review process, pushing approval times from a day to up to a month. In response, the company has begun pulling or blocking updates for apps that violate its self‑containment rules, sparking a standoff with the platforms that power the AI‑generated boom. Regulators are watching as the dispute highlights a clash between rapid AI development and existing gatekeeping frameworks. Lire la suite

A developer explored "vibe coding" by using three AI chatbots—Gemini, Claude, and ChatGPT—to create a single‑file web application called the Tome Reader. The app reads uploaded text aloud, highlights it in real time, and adds background music and sound effects triggered by specific words. Gemini handled the initial build, Claude refined trigger‑word handling, and ChatGPT reproduced the final version while struggling with a volume‑slider feature. The experiment highlighted each model's strengths and quirks, showing that a well‑crafted prompt can enable any of them to deliver a functional prototype. Lire la suite

LinkedIn is expanding its profile features to let users display verified proficiency in emerging AI coding tools, known as "vibe coding." Partnering with companies such as Replit, Lovable, Descript and Relay.app, the platform will allow these tool providers to assess users' skill levels and add them directly to profiles. The move aims to give recruiters a more reliable signal of candidates' AI tool usage while preserving existing profile signals. Lire la suite

Vibe coding lets non‑programmers create web apps by describing ideas to AI chatbots such as Gemini, ChatGPT, or Claude. Success hinges on knowing the strengths and limits of the chosen model, giving clear and exhaustive prompts, iterating through refinements, handling basic technical choices like HTML formatting, and staying flexible when bugs or scope limits appear. When progress stalls, starting a fresh chat can reset the workflow and improve outcomes. Lire la suite

A hands‑on experiment compared Google’s Gemini 3 Pro (a “thinking” model) with Gemini 2.5 Flash (a “fast” model) for vibe‑coding—a workflow that creates web projects through natural‑language prompts. Using the same project idea, a horror‑movie showcase, the author found the Pro model produced a more polished result with fewer manual steps, while the Flash model was quicker but required more specific prompting and frequent fixes. The test highlighted differences in speed, depth of reasoning, and user effort, offering insight for developers choosing between Gemini’s model tiers. Lire la suite

Vibe coding lets people without programming experience create functional apps by describing their vision to AI chatbots such as Gemini, ChatGPT, or Claude. The process hinges on picking the right model, offering clear and detailed prompts, and iterating through multiple revisions. While faster models may require more hands‑on editing, advanced “thinking” models handle larger portions of the code automatically. Users must also tackle basic technical choices, test for bugs, and stay open to restarting projects when needed. The approach empowers non‑technical creators to bring ideas to life with minimal code knowledge. Lire la suite

A hands‑on comparison of Google’s Gemini 3 Pro and Gemini 2.5 Flash models shows that the higher‑tier model delivers deeper reasoning and smoother code generation for vibe‑coding projects, while the faster model requires more manual prompting and frequent fixes. The experiment highlights trade‑offs between speed and depth, with Gemini 3 Pro generally producing more complete results with fewer iterations. Lire la suite

Vibe coding—using large language models to write software from prompts—offers faster development and broader accessibility, but it also introduces serious security risks. Studies show a significant portion of AI‑generated code contains serious flaws, and attackers can exploit poisoned code libraries to spread vulnerabilities. Experts stress that human oversight, strict code reviews, private sandboxed models, and Zero‑Trust access controls are essential to mitigate these threats while still benefiting from the efficiency of AI‑assisted development. Lire la suite

Vibe coding lets developers and non‑programmers create apps by describing functionality in plain language, with AI models generating the underlying code. Coined by Andrej Karpathy, the approach builds on recent advances in large language models and blurs the line between low‑code, no‑code, and traditional programming. While it speeds prototyping and opens software creation to a broader audience, experts warn that human oversight remains essential for security, performance, and maintainability. Lire la suite

Google's Gemini 3 model demonstrates a new ability to create playable retro‑style game demos using a conversational "vibe coding" approach. By guiding the model, users generated HTML versions of classics such as Chip's Challenge, Jezzball, an 8‑bit take on Tomba!, and a simplified Silent Hill experience. The AI provided step‑by‑step instructions for packaging games and compared favorably to the free ChatGPT version. Gemini's "Thinking" mode proved especially effective for deeper code generation, offering a potential tool for preserving and revisiting vintage games without the high cost of original hardware or software. Lire la suite

ChatGPT’s Agent, available to Plus and Pro subscribers, lets users assign complex or routine tasks for the AI to complete in the background. It can research topics, search for flights, write code, and generate presentations or spreadsheets, acting like a virtual assistant. While the feature speeds up workflows, it still requires human oversight for actions such as purchasing tickets, and it isn’t suited for simple, instant queries that are better handled by the standard chat mode. Lire la suite

A hands‑on comparison of three leading generative AI models—Google's Gemini 3 Pro, OpenAI's ChatGPT 5.1, and Anthropic's Claude Sonnet 4.5—was conducted by tasking each with building a web‑based Thumb Wars game. Gemini 3 Pro delivered the most complete and responsive code, automatically adding desktop keyboard controls and a 3‑D ring environment. ChatGPT 5.1 produced a functional prototype but required additional prompting for desktop support and lacked depth. Claude Sonnet 4.5 generated a playable demo with customization options but failed to implement promised keyboard controls. Overall, Gemini 3 Pro proved the fastest and most adaptable coder. Lire la suite

Developers are increasingly turning to AI‑generated code, dubbed “vibe coding,” to accelerate software creation. While the approach mirrors the efficiency of open‑source reuse, experts warn it introduces opaque code, potential vulnerabilities, and weakened accountability. Security firms highlight that AI models often draw on outdated or insecure codebases, making it hard to trace origins or audit outputs. A recent survey found that a third of security leaders report over 60 % of their code now originates from AI, yet fewer than one‑fifth have approved tools for such development. The emerging risk landscape calls for new safeguards and clearer governance. Lire la suite

Microsoft has rolled out an AI feature called Agent Mode for Word and Excel, powered by Microsoft 365 Copilot. The tool helps users create professional documents and spreadsheets by handling drafting, formula creation, and refinements, even for those with limited experience. An Office Agent in Copilot chat also assists in building PowerPoint decks from simple prompts. Agent Mode is available through the Frontier program for Microsoft 365 Copilot customers and Personal or Family subscribers, currently on the web with desktop versions coming soon. Lire la suite

Salesforce introduced Agentforce Vibes, an AI‑driven coding assistant that lets developers describe app requirements in natural language while the system generates code, integrates with existing Salesforce orgs, and enforces enterprise security and governance. The offering includes the Vibe Codey agent, built on an open‑source AI coding framework, and initially launches for free to existing users with usage limits tied to OpenAI’s GPT‑5 and Salesforce‑hosted Qwen 3.0 models. The move positions Salesforce deeper into the emerging vibe‑coding market as it balances innovation with cost and security considerations. Lire la suite

Ami Luttwak, chief technologist at Wiz, explains how the rapid adoption of artificial intelligence is expanding the attack surface for cybercriminals. While AI helps developers ship code faster, it also creates shortcuts and insecure implementations that attackers exploit. Luttwak highlights recent supply‑chain breaches, including the compromise of a chatbot startup and a popular JavaScript build system, where AI‑driven tools were used to harvest credentials and infiltrate corporate networks. He urges organizations to embed security from day one, adopt rigorous compliance standards, and rethink every layer of defense as AI continues to evolve. Lire la suite

AI‑assisted coding, often called vibe coding, lets developers describe functionality in plain language and have AI generate the code. Experts from Hostinger Horizons share five practical tips to get the most out of this approach: keep prompts short and specific, break tasks into smaller steps, ask the AI to suggest improvements, test frequently, and maintain human oversight. By following these guidelines, users can create more accurate, maintainable, and innovative applications while still leveraging the speed and accessibility of AI tools. Lire la suite

Rabbit has launched RabbitOS 2, a major software overhaul for the Rabbit R1 handheld AI device. The update adds a colorful card‑based interface, multimodal voice, text and image interactions, and a new "Creations" platform that lets users build tools, games and experiences through conversational "vibe coding". The update aims to close the gap between early expectations and the original experience, offering a more visual and intuitive way to access the R1’s capabilities while keeping the device’s low price and niche appeal. Lire la suite