<h2>Must know</h2>
<ul>
<li><strong>Microsoft’s February patch Tuesday addresses zero-days and AI vulnerabilities</strong>: Microsoft’s February 2026 Patch Tuesday fixed 59 vulnerabilities, including six actively exploited zero-days. Critically, it also patched remote code execution flaws in GitHub Copilot and multiple IDEs (VS Code, Visual Studio, JetBrains) stemming from AI prompt injection vulnerabilities. These AI-related flaws pose significant risks to developers due to potential access to sensitive credentials and infrastructure keys, as reported by Krebs on Security. <a href="https://app.daily.dev/feed-by-ids?id=zk4Rm0PBv&amp;id=BOPPPXswR&amp;id=i2YjgAdJA&amp;id=oUCth2teJ&amp;id=uxvhPzT2r&amp;id=zFYu8Wov8" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI agents introduce new security risks and governance challenges</strong>: The rapid deployment of autonomous AI agents, expected in 40% of enterprise applications by 2026, is creating significant security vulnerabilities and governance gaps. GitGuardian, which just raised $50M, highlights the crisis of non-human identities and secrets sprawl, especially as AI tools democratize code creation among non-technical users. Experts warn that AI agent ‘skills’ (markdown files providing context) are vulnerable to supply chain attacks and hidden commands, turning prompt injection into remote code execution if not properly secured with least privilege and task-based authorization. <a href="https://app.daily.dev/feed-by-ids?id=0rQwEgnbB&amp;id=QRBfTNO1G&amp;id=ZVqULJLz3&amp;id=gUwIhEnZO&amp;id=l8YCDT7qf&amp;id=mmQEWnSC4&amp;id=UhlBeqTGk&amp;id=1X0JlaofS&amp;id=48BFhb5Uw&amp;id=73Kh0XGFs" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Open-source GLM-5 language model released, claims parity with proprietary models</strong>: <a href="http://Z.ai" target="_blank" rel="noopener nofollow">Z.ai</a> released GLM-5, a new open-source large language model under an MIT License, claiming it matches proprietary models like Claude Opus 4.6 and GPT 5.3 Codex for long-horizon coding tasks. With 744B parameters and trained on 28.5T tokens, it integrates DeepSeek Sparse Attention and a novel asynchronous RL infrastructure. Modal is offering free access via an OpenAI-compatible API endpoint until the end of April. <a href="https://app.daily.dev/posts/bOspGaOnW" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>Kubernetes is now dominant for AI workloads, but culture is the new bottleneck</strong>: Kubernetes is now used in production by 82% of organizations for AI workloads, but cultural barriers like siloed teams and lack of cross-functional collaboration are preventing successful scaling of AI initiatives. <a href="https://app.daily.dev/posts/MrQz3g4zd" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Databricks launches Lakeflow for unified data engineering</strong>: Databricks introduced Lakeflow on Azure Databricks, a comprehensive solution for data ingestion, transformation, orchestration, and governance, featuring point-and-click ingestion, declarative pipelines, and AI-powered debugging. <a href="https://app.daily.dev/feed-by-ids?id=Bf1dK14FP&amp;id=DckO36Vaq&amp;id=cW6KgKgEK&amp;id=omSTWjEfj" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Vercel Flags enters public beta for feature flag management</strong>: Vercel Flags is now in public beta, offering native feature flag management with targeting rules, user segments, and environment controls, supporting Next.js, SvelteKit, and OpenFeature standard. <a href="https://app.daily.dev/posts/IfJOyXyyM" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Rust ecosystem continues to grow with AI coding assistant adoption</strong>: Rust’s ecosystem is expanding, attracting experienced developers for systems, CLI, and backend development, with 78% of Rust developers now using AI coding assistants like ChatGPT and GitHub Copilot. <a href="https://app.daily.dev/posts/LYg98DMrO" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Erlang/OTP 29.0-rc1 introduces native records and quantum-resistant SSH</strong>: Erlang/OTP 29.0-rc1 brings new features like native records, multi-valued comprehensions, and defaults to quantum-resistant <code>mlkem768x25519-sha256</code> key exchange for SSH, while dropping 32-bit Windows builds. <a href="https://app.daily.dev/posts/fTCktIrzF" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>ESLint 10.0.0 removes .eslintrc support, mandates flat config</strong>: ESLint 10.0.0 will remove support for <code>.eslintrc</code> configuration files, making flat config the sole option, improving config resolution for monorepos and fixing JSX reference tracking for React projects. <a href="https://app.daily.dev/posts/eQ0JUyyzE" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Amazon OpenSearch Serverless introduces Collection Groups for cost savings</strong>: Amazon OpenSearch Serverless now offers Collection Groups, allowing multiple collections to share compute resources (OCUs) while maintaining security isolation, primarily for cost reduction in multi-tenant workloads. <a href="https://app.daily.dev/posts/5AicsK9Gn" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>AI will eliminate 50% of white-collar jobs in 1-5 years, so buy my tools</strong>: An unnamed ‘industry leader’ is predicting that advanced AI models like GPT-5.3 Codex and Claude Opus 4.6 will eliminate 50% of entry-level white-collar jobs in law, finance, medicine, and software engineering within 1-5 years. The solution, apparently, is to immediately adopt paid AI tools and experiment daily. Because nothing says ‘imminent job loss’ like a sales pitch for your AI subscription. <a href="https://app.daily.dev/feed-by-ids?id=jQIyW7C3o&amp;id=lGyFb7XTB&amp;id=waJpUtc5V" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI agents are the future, but don’t ask about the security or the job losses</strong>: Google Cloud launched its GEAR program to help you build and deploy AI agents, competing with Microsoft, AWS, and OpenAI. They claim AI agents will bring 30-50% efficiency gains and enable ‘vibe coding’ for non-developers. Never mind the security vulnerabilities, employment displacement, or the fact that only 25% of enterprises have actually productionized their AI pilots. It’s a ‘watershed moment,’ just ignore the potential for a flood. <a href="https://app.daily.dev/feed-by-ids?id=9C4SjPZli&amp;id=gwxHevbDX&amp;id=o13HbD6JR&amp;id=r88T1MoRe&amp;id=vS730ONZY" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Recruitment is ‘broken’ and ‘Hunger Games’-like, but here’s how to fix it (for a fee)</strong>: Suzi Edwards-Alexander declares technical recruitment ‘fundamentally broken’ and a ‘race to the bottom’ driven by AI, comparing it to the Hunger Games. Her solution? Values-based structured interviews and treating interviewing as a skill. So, the system is broken, but don’t worry, she’s got the secret sauce to navigate it. Just make sure you practice your ‘language games’ and ‘culture-add’ talking points. <a href="https://app.daily.dev/posts/LD0ov49gc" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
