<h2>Must know</h2>
<ul>
<li><strong>NVIDIA’s Rubin platform redefines AI inference</strong>: NVIDIA’s new Rubin platform is a six-chip integrated system designed for AI inference, featuring NVLink 6 with 3.6 TB/s per GPU and HBM4 memory. It promises to run workloads with up to 4x fewer GPUs than Blackwell, reducing token costs significantly. This platform, shipping in late 2026, targets multi-agent systems, long-context reasoning, and real-time physical AI, as reported by NVIDIA. It’s a major leap for anyone building or deploying advanced AI. <a href="https://app.daily.dev/posts/828MqCIxP" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AWS European Sovereign Cloud is now live</strong>: AWS has launched its European Sovereign Cloud, with the first region in Brandenburg, Germany. This infrastructure is entirely within the EU, operated solely by EU residents, and physically and logically separate from other AWS regions. It offers comprehensive AWS services and aims to meet strict EU data residency and operational autonomy requirements, as announced by AWS. Canonical is a launch partner, providing Ubuntu and Ubuntu Pro. <a href="https://app.daily.dev/feed-by-ids?id=DTVSjPW7g&amp;id=B8P78hRXl" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Mandiant forces NTLMv1 deprecation with rainbow tables</strong>: Google-owned Mandiant has released comprehensive rainbow tables for Net-NTLMv1, making it trivial to crack credentials in under 12 hours with consumer-grade hardware. Their goal is to force organizations to abandon the 20-year-old insecure protocol, which is still prevalent in legacy systems. Microsoft has already removed NTLMv1 support from Windows Server 2025 and Windows 11, but Mandiant’s move highlights the urgent need for organizations to disable it, as detailed across multiple security news outlets. <a href="https://app.daily.dev/posts/2q2phjvWS" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>AI agents are changing software development roles</strong>: AI coding agents are shifting the bottleneck in software development from execution to problem definition, creating a new ‘Product Engineer’ role that blends technical skills with deep customer understanding and communication. <a href="https://app.daily.dev/feed-by-ids?id=ruSCTD0Fh&amp;id=sifgyKGyd" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>ClickHouse raises $400M, acquires Langfuse for AI analytics</strong>: ClickHouse secured $400 million in Series D funding and acquired Langfuse, an open-source LLM observability platform, aiming to create a unified analytics and debugging platform for enterprise AI workloads. <a href="https://app.daily.dev/posts/9QbO7XqRY" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>LLVM/Clang 22 adds support for new Intel, Arm, and Ampere CPUs</strong>: LLVM/Clang 22 feature development is complete, bringing support for Intel Nova Lake, Arm C1, and Ampere1C processors, along with C2y language features and Zen 4 optimizations, with the stable release expected by late February. <a href="https://app.daily.dev/posts/DFbxDQnoL" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Fedora Games Lab switches to KDE Plasma</strong>: Fedora Games Lab is moving to KDE Plasma to modernize its Linux gaming distribution, leveraging Wayland, PipeWire, HDR, and VRR support, and improving documentation and handheld device compatibility. <a href="https://app.daily.dev/posts/55802yXlT" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Node.js releases address multiple CVEs</strong>: Node.js versions 20.20.0 LTS and 25.3.0 were released, patching several security vulnerabilities including a critical DoS flaw (CVE-2025-59466) affecting async_hooks and AsyncLocalStorage. <a href="https://app.daily.dev/feed-by-ids?id=DwsoMgMKH&amp;id=EyLXGRUe2&amp;id=WX4br4Zyw&amp;id=Yy5NFIIOM&amp;id=Zr0Dv9dwZ" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Docker security hardening checklist released</strong>: A new security-focused checklist provides actionable items for hardening Docker containers, images, and runtime environments, crucial for securing containerized applications and infrastructure. <a href="https://app.daily.dev/posts/BjYgCt53T" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Pulumi adds native support for Terraform and HCL</strong>: Pulumi now natively supports Terraform and HCL, allowing direct HCL execution and state management within Pulumi Cloud, easing migration and enabling hybrid infrastructure-as-code workflows. <a href="https://app.daily.dev/posts/brDXTtnxJ" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>AI coding tools: the emperor has no clothes (or working code)</strong>: Cursor, the AI-powered code editor, proudly announced its agents built a web browser in a week, generating over a million lines of code. Turns out, the code doesn’t even compile, with 34 errors and 94 warnings, and no working demo. It seems ‘meaningful progress’ now means ‘impressive screenshots’ and ‘non-functional AI-generated output.’ Maybe we should stick to human developers for things that actually need to, you know, <em>work</em>. <a href="https://app.daily.dev/feed-by-ids?id=GPLZOgOgP&amp;id=V3KHuMgUU" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI is destroying developer skills, but don’t worry, it’s ‘progress’</strong>: Apparently, AI coding assistants are so good they’re eroding fundamental developer skills like reading documentation and critical thinking. Tailwind CSS even laid off 75% of its engineers because developers stopped visiting their docs. The real risk isn’t job loss, it’s the degradation of problem-solving abilities across the industry, turning us all into prompt monkeys. But hey, at least we can generate more ‘slop’ faster, right? <a href="https://app.daily.dev/feed-by-ids?id=bNykMIpel&amp;id=DdiXGTqPN" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Small language models: the new ‘disruptive innovation’ that will save us from LLM ‘liabilities’</strong>: Enterprises are suddenly ‘favoring’ Small Language Models (SLMs) because LLMs are too expensive and open-ended. Now, your 100-billion-parameter behemoth is a ‘liability,’ and 7-million-parameter models are ‘outperforming’ frontier LLMs on ‘constrained tasks.’ It’s almost like the industry just discovered that a specialized tool is better for a specialized job than a general-purpose one. Who knew? Get ready for the next wave of hype, where SLMs are the ‘future-defining shift’ until the next big thing comes along. <a href="https://app.daily.dev/feed-by-ids?id=HX43N3MkE&amp;id=J1BKrq2u6" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
