<h2>Must know</h2>
<ul>
<li><strong>AI coding assistants: productivity vs. pitfalls</strong>: AI coding tools like GitHub Copilot are showing mixed results. While some report significant productivity boosts (up to 55%), others find increased security vulnerabilities (322% more) and lower code stability. Experienced developers often see negligible gains due to the time spent verifying AI output, as highlighted by METR and Ruby on Rails creator David Heinemeier Hansson. The consensus is that AI agents act more like junior engineers, requiring precise specifications and careful oversight to avoid introducing technical debt and quality issues. <a href="https://app.daily.dev/feed-by-ids?id=0fzwUL3Um&amp;id=15Mjf12HP&amp;id=2IKuR9DJZ&amp;id=3DI7vnha4&amp;id=3sM5oeueQ&amp;id=AJzG7x8FZ&amp;id=AS53pzdrZ&amp;id=ATZPMmQn8&amp;id=DdiXGTqPN&amp;id=EqMZfdLoq" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Microsoft Intune enforces mandatory app updates</strong>: Microsoft Intune will soon require all iOS and Android enterprise apps to update to the latest SDK versions by January 19, 2026. Failure to comply will result in apps being blocked from launching, affecting both Microsoft and third-party applications. This move, which includes new security features like PIN enforcement and biometric authentication, is a non-optional enforcement, effectively acting as a ‘kill switch’ for non-compliant apps, according to reporting. <a href="https://app.daily.dev/posts/PNEoUht10" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AMD Zen processors vulnerable to StackWarp attack</strong>: A new hardware vulnerability, StackWarp (CVE-2025-29943), affects AMD Zen processors (generations 1-5), including EPYC 7003, 8004, 9004, and 9005 series. This flaw allows privileged attackers to execute arbitrary code within SEV-SNP protected confidential virtual machines by exploiting an undocumented control bit. AMD has issued microcode updates, with more scheduled for April 2026, and advises disabling hyperthreading on critical systems as a temporary mitigation. <a href="https://app.daily.dev/posts/9DURDnZX3" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>GitLab launches Duo Agent Platform for AI orchestration</strong>: GitLab 18.8 introduces the Duo Agent Platform, enabling orchestration of multiple AI agents across the software development lifecycle, including a Planner Agent and a Security Analyst Agent for automated vulnerability triage. <a href="https://app.daily.dev/posts/R9WyVAVPv" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Intel updates LLM-Scaler-Omni for Arc Graphics</strong>: Intel released an update to its LLM-Scaler-Omni, enhancing support for ComfyUI and SGLang on Intel Arc Graphics, improving multi-modal generative AI capabilities for image, voice, and video generation. <a href="https://app.daily.dev/posts/XBX3pzO5M" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Microsoft open-sources XAML Studio</strong>: Microsoft has open-sourced XAML Studio under the .NET Foundation, making its lightweight prototyping tool for XAML UI development available on GitHub for community contributions and deeper integration with WinUI 3. <a href="https://app.daily.dev/posts/hUJrXK08w" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Rust expands platform support with Socket</strong>: Socket now offers General Availability for Rust and Cargo support, providing dependency analysis, SBOM generation, and supply chain visibility for Rust projects, addressing risks beyond memory safety. <a href="https://app.daily.dev/posts/O8W5FyqYs" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>New language models: GLM-4.7-Flash and Differential Transformer V2</strong>: GLM-4.7-Flash, a 30B-A3B Mixture of Experts model, and Differential Transformer V2, with a revamped attention mechanism, are pushing boundaries in efficient and stable language model architectures. <a href="https://app.daily.dev/feed-by-ids?id=0sHlXeafO&amp;id=a35LKDHC2" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Microsoft patches Windows 11 shutdown bug</strong>: Microsoft issued an emergency out-of-band update (KB5077797) for Windows 11 23H2 to fix a critical bug from the January Patch Tuesday that prevented proper shutdown, restart, or hibernation, also addressing Remote Desktop authentication issues. <a href="https://app.daily.dev/feed-by-ids?id=u2UYa6XNM&amp;id=KR0gCuX2r" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>PHP 8.4 introduces breaking change in http_build_query()</strong>: PHP 8.4’s backward-incompatible change in <code>http_build_query()</code> now causes serialization issues with backed enums and errors on non-backed enums, breaking legacy code and prompting new URI toolkit solutions for compatibility. <a href="https://app.daily.dev/posts/FEB00NeSK" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>AI website builders: shipping or just hype?</strong>: Apparently, we now have ‘Agentic AI Website Builders’ that ‘actually ship.’ Tools like Replit Agent and v0 promise to create full-stack apps with autonomous testing and one-click deployment. Because, you know, the last thing we needed was more AI-generated boilerplate code without a human in sight to ensure it actually makes sense or is maintainable. What could possibly go wrong? <a href="https://app.daily.dev/posts/e3oUC0g8q" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI-powered coding tools for rapid app development</strong>: <a href="http://Pythagora.ai" target="_blank" rel="noopener nofollow">Pythagora.ai</a> claims to generate full-stack applications from text prompts in hours, while a new book teaches ‘vibe coding’ for games using AI assistants like Gemini and Grok. It seems the dream of non-developers creating functional games and apps with minimal effort is alive and well. Just don’t ask about the underlying architecture, security, or long-term maintainability—those details are for the ‘old-fashioned’ human engineers. <a href="https://app.daily.dev/feed-by-ids?id=LklGkgpse&amp;id=LpF1NSIQw" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI in customer onboarding and lead generation: zero hallucination claims</strong>: Companies like Remote and Clarm are touting AI automation for customer onboarding and lead generation, with bold claims of ‘zero-hallucination AI responses.’ Remote uses LangChain to standardize HR data, while Clarm leverages Vespa for GitHub star-to-revenue conversion. It’s almost as if they’ve solved the fundamental problems of LLMs, or perhaps, they’re just really good at marketing their ‘hybrid AI approach’ to bypass those pesky hallucination risks. We’ll see how that holds up in the real world. <a href="https://app.daily.dev/feed-by-ids?id=LVS7zKWbX&amp;id=x5GAIQ3mP" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
