<h2>Must know</h2>
<ul>
<li><strong>Kubernetes dashboard radar offers real-time visibility</strong>: A new open-source Kubernetes dashboard called Radar has been released, built as a single Go binary with an embedded React frontend. It provides real-time cluster visibility through topology graphs, event timelines, and Helm management without requiring cloud logins or agents. Radar natively supports GitOps tools like Argo CD and Flux, running locally using existing kubeconfig credentials, addressing limitations of existing tools like Lens and k9s. <a href="https://app.daily.dev/feed-by-ids?id=DiEunA7dc&amp;id=a3wJg58hN" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Cloud native ecosystem sees major shifts</strong>: The cloud native ecosystem is undergoing significant changes, with the Ingress NGINX project announcing its retirement and the Kubernetes Dashboard project being archived. On a positive note, Cluster API v1.12 was released, including in-place updates, and a new Checkpoint/Restore Working Group has been formed. These developments highlight the dynamic nature of cloud-native infrastructure, as reported by CloudNative.Now. <a href="https://app.daily.dev/posts/EdFbm39PB" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI security and privacy concerns escalate</strong>: Researchers found over 175,000 Ollama AI hosts exposed online without safety guardrails, enabling potential resource hijacking and remote attacks. SpaceX updated Starlink’s privacy policy to use customer data for AI training, coinciding with talks to merge with xAI. Additionally, a critical 1-click RCE vulnerability was discovered and patched in OpenClaw, an AI assistant used by over 100,000 developers, highlighting ongoing security risks in the AI landscape. <a href="https://app.daily.dev/feed-by-ids?id=2SdHHyeRJ&amp;id=YUSdAM3D6&amp;id=d76FYhLFT&amp;id=iEIaVGrhd&amp;id=yXHHUa4yo&amp;id=4nlzVZMEL&amp;id=ALGXJz29U&amp;id=fuL6WfKHt&amp;id=oAoGgXAAn&amp;id=zgJdriYuf" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>Wix introduces xEngineer model for AI-driven development</strong>: Wix launched the ‘xEngineer’ model, integrating AI into engineering roles to broaden ownership across the development lifecycle, with AI supporting tasks like signal interpretation and failure triage. <a href="https://app.daily.dev/posts/1ESLU5xBs" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI’s evolving impact on software development</strong>: AI tools now mediate between developers and documentation, with 84% of developers using them daily, requiring documentation to serve both human and AI audiences. <a href="https://app.daily.dev/feed-by-ids?id=3L44pW4ot&amp;id=5ZJWT4v6q&amp;id=JVTlCrLRm&amp;id=h7CGoaIP0&amp;id=zVcgznCs0" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>MLflow integrates new evaluation metrics for LLMs</strong>: MLflow 3.9.0 now integrates over 50 evaluation metrics from DeepEval, RAGAS, and Arize Phoenix frameworks, providing a unified API to evaluate LLM agents and RAG systems. <a href="https://app.daily.dev/posts/zVcgznCs0" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>NVIDIA NVFP4 quantization released by Red Hat</strong>: Red Hat has released NVFP4-quantized versions of popular LLMs, leveraging NVIDIA’s 4-bit floating-point format for efficient deployment with near-baseline accuracy on Blackwell GPUs. <a href="https://app.daily.dev/posts/OPe22PWoI" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>DataDog releases pg_tracing for PostgreSQL</strong>: DataDog developed <code>pg_tracing</code>, a PostgreSQL extension for server-side distributed tracing that generates spans for database queries and operations, supporting PostgreSQL 14, 15, and 16. <a href="https://app.daily.dev/posts/Xfr0u4GSj" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>WebAssembly celebrates 10 years, expands beyond browser</strong>: WebAssembly (WASM) marked its 10th anniversary, evolving from a browser performance solution to a foundational technology for edge platforms, serverless runtimes, and embedded systems, promising lightweight, composable modules. <a href="https://app.daily.dev/posts/XjOBMuNCm" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Linux kernel 6.19 to include Uniwill/TUXEDO driver</strong>: The upcoming Linux 6.19 kernel will integrate a new Uniwill laptop platform driver, enabling cTGP graphics power settings and other platform-specific controls for Uniwill-manufactured and TUXEDO Computers models. <a href="https://app.daily.dev/posts/mzBRRS79C" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>AI code generation: 55% faster, 91% longer reviews</strong>: Apparently, AI code generation makes you 55% faster at writing code, but then your review times jump by 91% because, surprise, AI code needs a lot more babysitting. Microsoft’s even pushing out emergency patches more often, claiming 30% of its code is AI-generated. So, we’re just shifting the bottleneck from writing bad code to reviewing bad AI-generated code. Progress! <a href="https://app.daily.dev/feed-by-ids?id=CUMsxmpoK&amp;id=H1aWwdchu&amp;id=KYToiOjQs&amp;id=TCb4dVj23" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI’s limited impact on product development</strong>: Some ‘experts’ are still trying to convince us that AI hasn’t fundamentally changed product development, just added new tools. They claim 10x-100x productivity gains lack validation and that human developers are still essential. It’s almost like they’re trying to justify their own existence while tiny teams with advanced AI tools are ‘outcompeting organizations thousands of times their size.’ Keep telling yourself that, chief. <a href="https://app.daily.dev/feed-by-ids?id=R5VZVKE0f&amp;id=WZuX1OWTx" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Indonesia lifts Grok ban after ‘service improvements’</strong>: Indonesia, Malaysia, and the Philippines have conditionally lifted their bans on xAI’s Grok chatbot after it generated millions of nonconsensual sexualized images. xAI ‘committed to service improvements’ and restricted image generation to paying subscribers. Because nothing says ‘responsible AI’ like a quick fix after a massive scandal, and then charging people for the privilege of using your now-less-problematic tool. California’s AG is still investigating, but hey, at least the ban’s lifted! <a href="https://app.daily.dev/posts/xKtMniybW" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
