<h2>Must know</h2>
<ul>
<li><strong>AI coding assistants are making software development slower and buggier</strong>: Turns out, those shiny AI coding assistants aren’t the productivity panacea we were promised. A 2025 study found developers using AI took 19% longer to complete tasks, and code review times on high-AI-usage teams shot up by 91%. Plus, 45% of AI-generated code comes with security vulnerabilities. It seems treating AI like a junior developer needing constant supervision and detailed specs is the only way to avoid a ‘zig-zag problem’ of endless iterations, as reported by various tech outlets. <a href="https://app.daily.dev/feed-by-ids?id=pYjmmMIkn&amp;id=MVxk3C8r1&amp;id=2tsKdG55u" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>TSMC’s capital expenditure bottleneck is choking AI chip supply</strong>: TSMC’s flat capital expenditures since 2021, despite a surge in hyperscaler demand post-ChatGPT, have created a massive AI chip supply bottleneck. This imbalance is projected to cost hyperscalers billions in lost revenue and persist until at least 2028. The only real solution? More competition from Samsung or Intel to force TSMC’s hand, according to industry analysts, who suggest hyperscalers diversify their supply chains now. <a href="https://app.daily.dev/posts/pPxEIulIE" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Microsoft launches Maia 200 AI inference accelerator</strong>: Microsoft just dropped the Maia 200, a custom AI inference accelerator built on TSMC’s 3nm process. It boasts over 10 petaFLOPS in FP4 precision and 216GB HBM3e memory, claiming 30% better performance per dollar than current hardware. This beast is slated for GPT-5.2 models and Microsoft 365 Copilot, with initial deployment in Microsoft’s US Central datacenter, as announced by Microsoft. <a href="https://app.daily.dev/posts/HBgJElvHn" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>Node.js 25.5.0 brings security and performance updates</strong>: Node.js 25.5.0 is out, upgrading root certificates to NSS 3.119, adding LIEF for Single Executable Application (SEA) generation, and enhancing <code>fs.watch</code> with an ‘ignore’ option. <a href="https://app.daily.dev/posts/BrQRGrBKE" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Rust core team member creates new AI-assisted systems language</strong>: Steve Klabnik, a Rust core team member, unveiled Rue, a new memory-safe systems language that ditches the borrow checker for ‘inout’ parameters and was largely developed with Anthropic’s Claude AI. <a href="https://app.daily.dev/posts/scnD6visU" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Docker Sandboxes isolates AI coding agents with microVMs</strong>: Docker Sandboxes now uses microVMs to isolate AI coding agents like Claude Code and Gemini CLI, allowing them to install packages and modify files without impacting the host system, currently on macOS with Windows and Linux support planned. <a href="https://app.daily.dev/posts/6YdOupFK7" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>OpenAI upgrades ChatGPT code execution with Bash and package installs</strong>: ChatGPT’s code execution feature now supports Bash commands and package installation via pip and npm through an internal proxy, expanding its utility for all users, including free-tier. <a href="https://app.daily.dev/posts/7V1v7MpYn" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Google launches FunctionGemma for mobile and edge AI agents</strong>: Google released FunctionGemma, a 270M parameter model optimized for function calling on mobile and edge devices, translating natural language into structured API calls with 85% accuracy. <a href="https://app.daily.dev/posts/DFyxBQQFb" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>NVIDIA MPS boosts small LLM GPU throughput</strong>: NVIDIA’s Multi-Process Service (MPS) can increase GPU throughput by 50-100% for small language models (under 3B parameters) with short contexts, by allowing multiple inference processes to share GPU resources and overlap kernel operations. <a href="https://app.daily.dev/posts/0yNJa1gH5" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Uncloud gains traction as a simpler Kubernetes alternative</strong>: Uncloud, an open-source deployment platform built on Docker, Caddy, and WireGuard, is gaining momentum among small teams as a simpler Kubernetes alternative, offering single-command deployments and nearing a stability milestone. <a href="https://app.daily.dev/posts/FrS1fa85P" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>Cursor AI agents built a browser ‘from scratch’ (spoiler: they didn’t)</strong>: Cursor’s CEO proudly announced their AI agents built a ‘fully functional web browser’ in a week, ‘from scratch,’ generating over 3 million lines of code. Independent developers, however, found the codebase barely compiles, relies heavily on existing projects like Servo and QuickJS, and performs ‘abysmally.’ So, ‘from scratch’ apparently means ‘cobbled together from open-source projects and still doesn’t work.’ Impressive marketing, though. <a href="https://app.daily.dev/posts/DovPZUpOP" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI is making open source unsustainable, apparently</strong>: Tailwind CSS, a popular open-source project, saw an 80% revenue drop and laid off 75% of its staff, blaming developers using AI tools instead of official docs. Now, Frontend Masters is calling for corporate sponsorship, claiming AI is causing an ‘existential crisis’ for open source. Because, clearly, the problem isn’t that companies profit immensely from open source without contributing, it’s that AI is making developers too efficient to click on sponsored links. <a href="https://app.daily.dev/feed-by-ids?id=3KjVRqoNX&amp;id=YRC8HFFhU" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI chatbots are now citing fake encyclopedias</strong>: Both ChatGPT (GPT-5.2) and Anthropic’s Claude have been caught citing ‘Grokipedia,’ an AI-generated encyclopedia by Elon Musk, in their responses. Grokipedia is known for its conservative bias and inaccuracies, especially on historical events. So, our advanced AI models are now confidently hallucinating with sources, and OpenAI’s response is basically ‘we use a wide range of public sources.’ Transparency, folks, transparency. <a href="https://app.daily.dev/posts/SDpas3mop" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
