<h2>Must know</h2>
<ul>
<li><strong>AI coding assistants are a double-edged sword</strong>: AI coding assistants, while boosting productivity in code writing, demand significantly more upfront architectural planning and specification work. A 2025 study found developers with AI assistants took 19% longer to complete tasks, despite feeling faster. Code review times on high-AI-usage teams have increased by 91%, and 45% of AI-generated code contains security vulnerabilities, according to a report from TechCrunch. This suggests a hidden ‘architecture tax’ that can offset perceived productivity gains. <a href="https://app.daily.dev/posts/pYjmmMIkn" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>OpenAI and Google expand AI code execution capabilities</strong>: OpenAI has upgraded ChatGPT’s code execution feature to support Bash commands and multiple programming languages, including Python, JavaScript, and C++, along with package installation via pip and npm. This significantly expands its utility for developers. Meanwhile, Google released FunctionGemma, a 270M parameter model optimized for function calling on mobile and edge devices, enabling AI agents to perform on-device task automation with increased accuracy, as reported by Google’s official channels. <a href="https://app.daily.dev/feed-by-ids?id=7V1v7MpYn&amp;id=DFyxBQQFb" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>TSMC’s flat CapEx creates AI chip supply bottleneck</strong>: TSMC’s decision to maintain flat capital expenditures from 2021 to 2024, despite a surge in demand post-ChatGPT, has created a significant supply bottleneck for AI chips. This imbalance is costing hyperscalers billions in lost AI revenue, and shortages are projected to persist through at least 2028. Analysts suggest hyperscalers should prioritize supply diversification to mitigate this risk, as reported by The Information. <a href="https://app.daily.dev/posts/pPxEIulIE" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>Rust core team member builds new language with AI</strong>: Steve Klabnik, a Rust core team member, developed a new memory-safe systems programming language called Rue in two weeks, with substantial assistance from Anthropic’s Claude AI, generating approximately 70,000 lines of compiler code. <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</strong>: Docker Sandboxes now offers microVM-based isolation for AI coding agents like Claude Code and Gemini CLI, allowing them to install packages and modify files without impacting the host system, currently available 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>AI-generated software tests are often ineffective</strong>: Automated tests generated by large language models (LLMs) are criticized for creating a false sense of security, as they often fail to validate actual production behavior and can lead to brittle tests that break with internal refactoring. <a href="https://app.daily.dev/posts/wLM9iDorf" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Linux 6.19 rolls back scheduler changes</strong>: The Linux 6.19 kernel’s initial test versions experienced performance regressions with the NEXT_BUDDY feature for the EEVDF scheduler, leading to its disabling in the rc7 release candidate. <a href="https://app.daily.dev/feed-by-ids?id=85ivk605B&amp;id=mNRQJiXzS" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AMDGPU Linux driver adds HDMI VRR and ALLM</strong>: AMDGPU Linux kernel driver patches now introduce HDMI Variable Rate Refresh (VRR) and Auto Low Latency Mode (ALLM) for gaming on HDMI displays, aiming for parity with FreeSync capabilities. <a href="https://app.daily.dev/posts/bsF1QAXcQ" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Uncloud gains momentum as Kubernetes alternative</strong>: Uncloud, an open-source deployment platform using Docker, Caddy, and WireGuard, is gaining traction among small teams as a simpler Kubernetes alternative, supporting single-command deployments from source to production. <a href="https://app.daily.dev/posts/FrS1fa85P" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Node.js 25.5.0 released with security and feature updates</strong>: Node.js 25.5.0 includes an upgrade to NSS 3.119 root certificates, adds LIEF for Single Executable Application (SEA) generation, and introduces an ‘ignore’ option for fs.watch. <a href="https://app.daily.dev/posts/BrQRGrBKE" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>Cursor AI agents build a browser ‘from scratch’ (spoiler: they didn’t)</strong>: Cursor’s CEO proudly announced their AI agents built a fully functional web browser ‘from scratch’ in a week, generating over 3 million lines of code. Independent developers, however, quickly found the browser barely compiles, doesn’t work, and relies heavily on existing projects like Servo and QuickJS. So much for ‘from scratch’—seems the AI is better at generating marketing hype than functional software, as reported by TechCrunch. <a href="https://app.daily.dev/posts/DovPZUpOP" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI coding assistants are privatizing knowledge (and making you pay for it)</strong>: Apparently, LLM-powered coding assistants are increasingly trained on proprietary user data, leading to a ‘knowledge hoarding’ scenario where critical programming knowledge becomes paywalled. The author predicts market consolidation and employer-mandated LLM subscriptions. Because, you know, the open-source community just loves paying for what used to be free, as reported by The Register. <a href="https://app.daily.dev/posts/p61M36Xco" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI bubble to burst, says Norway (but still wants age limits for chatbots)</strong>: Norway’s Technology Council warns the ‘AI bubble’ might burst soon as company spending outpaces earnings, yet simultaneously suggests establishing age limits for AI services. Because nothing says ‘imminent financial collapse’ like regulating who can chat with a bot. They also want to control rocket launches and secure critical minerals, just in case the AI apocalypse doesn’t get us first, as reported by Reuters. <a href="https://app.daily.dev/posts/jDJWUdYgk" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
