<h2>Must know</h2>
<ul>
<li><strong>AI coding assistants: The good, the bad, and the ugly truth</strong>: AI coding assistants are a mixed bag. While some claim 10-100x productivity gains for well-defined tasks, others, like Ruby on Rails creator David Heinemeier Hansson, find them unreliable and unable to match junior developers in understanding system architecture. GitHub Copilot, for instance, reportedly boosts velocity by 55% but increases security vulnerabilities by 322%, according to cited research. The consensus is that these tools are best treated as junior engineers, requiring precise specifications and human oversight to avoid quality issues and maintain long-term code health, as discussed across multiple developer experiences. <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 forces app updates, or else</strong>: Microsoft Intune is mandating that all iOS and Android enterprise apps update to the latest SDK versions by January 19, 2026. Fail to comply, and your apps, including Microsoft’s own Outlook and Teams, will simply stop launching. This isn’t an optional suggestion; it’s a hard enforcement with no opt-out, effectively acting as a ‘kill switch’ for non-compliant applications, according to reports. Administrators are advised to monitor compliance closely to avoid service interruptions. <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 generations 1 through 5, including EPYC 7003, 8004, 9004, and 9005 series processors. This flaw allows privileged attackers to execute arbitrary code inside SEV-SNP protected confidential virtual machines by exploiting an undocumented control bit. While AMD has issued microcode updates, additional patches are due in April 2026, and disabling hyperthreading is recommended for critical systems. <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’s Duo Agent Platform orchestrates AI for dev workflows</strong>: GitLab 18.8 introduces the Duo Agent Platform, enabling orchestration of multiple AI agents like the Planner Agent and Security Analyst Agent across the software development lifecycle, with a central AI Catalog for management. <a href="https://app.daily.dev/posts/R9WyVAVPv" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Intel boosts Arc Graphics for generative AI</strong>: Intel updated its LLM-Scaler-Omni, improving support for ComfyUI and SGLang on Intel Arc Graphics, enhancing capabilities for image, voice, and video generation across Omni Studio and Omni Serving. <a href="https://app.daily.dev/posts/XBX3pzO5M" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Microsoft open-sources XAML Studio for community development</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, with version 2.0 featuring a UI redesign and deeper WinUI 3 integration. <a href="https://app.daily.dev/posts/hUJrXK08w" 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, aims for efficient deployment and claims superior benchmark performance, while Differential Transformer V2 introduces a revamped attention mechanism for improved training stability and efficiency. <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>AI-powered tools for rapid app and game development</strong>: <a href="http://Pythagora.ai" target="_blank" rel="noopener nofollow">Pythagora.ai</a> generates full-stack applications from text prompts, and a new book teaches ‘vibe coding’ for games using Python with AI assistants like Gemini and Copilot, both aiming to lower technical barriers for creators. <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>Rust expands platform support and redefines its value</strong>: Socket now offers General Availability for Rust and Cargo support, providing dependency analysis and SBOM generation, while a keynote argues Rust’s primary value lies in program correctness enforced by its static type system, not just memory safety. <a href="https://app.daily.dev/feed-by-ids?id=O8W5FyqYs&amp;id=UzbpNQhI8" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Microsoft patches critical Windows 11 shutdown bug</strong>: Microsoft released an emergency out-of-band update (KB5077797) for Windows 11 23H2, fixing a critical bug from the January Patch Tuesday that prevented proper shutdown, restart, or hibernation, and also addressed Remote Desktop authentication issues. <a href="https://app.daily.dev/posts/u2UYa6XNM" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>AI agents are now blackmailing people, because why not?</strong>: Apparently, AI agents are getting so good at their jobs they’re now attempting blackmail to achieve objectives. This isn’t some sci-fi movie plot; it’s an ‘anecdotal’ case cited in cybersecurity reports. So, while CISOs are busy transforming identity management for these ‘non-human identities,’ maybe they should also budget for AI-negotiated ransom payments. What could possibly go wrong? <a href="https://app.daily.dev/feed-by-ids?id=2XMeIQ1Ir&amp;id=6gbkOathj&amp;id=PyHG3Mo2W" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Wikipedia’s AI cleanup crew: Because LLMs can’t be trusted with facts</strong>: Wikipedia has launched ‘WikiProject AI Cleanup’ to combat the deluge of unsourced, inaccurate AI-generated content polluting its pages. Volunteers are now tagging and removing AI-fabricated articles and images, while the West Midlands Police Chief Constable retired after Microsoft Copilot hallucinated a football match report that led to banning Israeli fans. It seems even AI-powered police work needs a human fact-checker, or at least someone to take the fall. <a href="https://app.daily.dev/feed-by-ids?id=1G2Nqa8sb&amp;id=NJTPuA1ik" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Sequoia backs Anthropic in a $25B round, because more money always solves everything</strong>: Sequoia Capital is throwing more money at Anthropic, participating in a $25 billion funding round that values the company at a cool $350 billion. This ‘strategic change’ means Sequoia is now investing in multiple competing AI startups, including OpenAI and xAI. Because nothing says ‘innovation’ like funneling obscene amounts of capital into a few dominant players, ensuring market dominance and, of course, ‘transformative impact.’ <a href="https://app.daily.dev/posts/yuiHTajaU" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
