Blog
Notes from real projects on modern web development with Next.js, React and TypeScript, AI integrations with LLMs and RAG, and technical SEO.
425 posts · Page 18 of 21
OpenAI is turning ChatGPT from a chatbot into a full-fledged platform. With the new app directory and the Apps SDK, services like Apple Music and Figma are being integrated, and a whole new era is beginning for developers.
DRAM prices rose 172% in 2025, and in 2026 SD cards are coming back to phones. Behind the global memory crisis are HBM, AI and 'storage inflation'. But a storm that began in late 2023 and peaked in 2025 has turned this equation upside down.
Meta is working on a new AI project codenamed 'Mango' that will rival Google's Nano Banana model. Due in 2026, the model aims to revolutionize both image and video generation.
YouTube has started taking tough measures against AI-generated fake movie trailers that mislead viewers. The shutdown of giant channels with millions of subscribers, such as Screen Culture and KH Studio, signals a new era on the platform.
OpenAI announced that Apple Music will be integrated into ChatGPT. Soon you'll be able to build playlists in seconds just by talking to AI. This integration is set to fundamentally change our music listening habits.
Google is making fundamental changes to the Gemini app. The new interface spotted in the Android beta, a categorized 'My Stuff' section, AI-powered shopping tracking and a redesigned text input field are transforming the user experience from top to bottom.
Mobvoi's TicNote Pods are next-gen earbuds that use 4G eSIM and Shadow AI to record and transcribe conversations without needing a phone. With claims of live transcripts and summaries, they focus on scenarios such as meetings, interviews and lecture notes.
Cursor's Debug Mode (an agent loop driven by runtime data) and its in-IDE browser approach move debugging in web projects from speculation to an evidence-based process. In this post, I explain how it works technically and its practical benefits for beginners.
According to shared information, OpenAI aims to strengthen spreadsheet/presentation generation, vision, coding and long-context performance with GPT-5.2. The company also highlights suitability for agentic workflows and claims fewer hallucinations.
What is CVE-2025-55182, the critical RCE vulnerability affecting projects that use React 19 and the Next.js App Router, who does it affect and how can you secure your project step by step? Here is a practical, actionable guide.
While ChatGPT maintains its global leadership, its growth rate is slowing noticeably; Google Gemini, meanwhile, is rapidly gaining momentum in metrics such as downloads, user growth and time spent in the app.
A new MIT study finds that AI could replace 11.7% of the US workforce, or roughly 20 million workers. The study projects that automation could lead to a $1.2 trillion shift in wages.
The balance of power in AI is shifting. OpenAI CEO Sam Altman issued a “code red” call to employees, asking them to turn their full focus to improving ChatGPT. The rise of Google and Anthropic is forcing OpenAI to change strategy.
Ad infrastructure code found in the latest beta of OpenAI's ChatGPT Android app points to a historic change in the platform's revenue model. What do SearchAdsCarousel and AdTarget reveal?
DeepSeek's new model Math-V2 reaches gold-medal level at the International Mathematical Olympiad, challenging Google and OpenAI. And it is fully open source.
Black Forest Labs has announced FLUX.2, its new text-to-image model. Thanks to a collaboration with NVIDIA, the model breaks through the 90 GB VRAM barrier, comes to consumer-grade PCs and has become Nano Banana Pro's strongest rival.
OpenAI has announced 'Shopping Research', a ChatGPT feature that creates personalized shopping guides. The tool runs detailed product comparisons based on users' budgets and preferences, opening a new era in e-commerce.
Anthropic has launched its new flagship model Opus 4.5. Breaking records in coding, agent systems and long-context operations, the model outperforms Gemini 3 and GPT-5.1. I looked at every detail of this new model, which shifts the balance of power in AI, along with a comparative analysis.
Designed specifically for Large Language Models (LLMs), TOON is a readable, efficient data format that uses 30-60% fewer tokens than JSON.
Google has introduced Nano Banana Pro (Gemini 3 Pro Image), its new image generation model. With text rendering, multi-image composition and C2PA-backed safety features, the model gives content creators studio-quality control.
Nvidia beat expectations by raising revenue to $57 billion in a quarter of record data center GPU sales. While saying cloud GPUs are sold out, the company expects AI investment to continue at full speed.