<h2>Must know</h2>
<ul>
<li><strong>OpenAI introduces ads to free and Go tiers</strong>: OpenAI is rolling out targeted advertising in the free and Go tiers of ChatGPT for US users, a move driven by significant financial losses and low paid subscriber numbers. Ads will appear at the bottom of conversations, clearly labeled, and OpenAI assures they won’t influence chatbot responses or involve selling user data, as reported by multiple sources. This marks a strategic shift from CEO Sam Altman’s previous stance that ads would be a “last resort” for monetization. <a href="https://app.daily.dev/posts/DQh9IAXNl" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Mandiant releases Net-NTLMv1 rainbow tables</strong>: Mandiant has publicly released rainbow tables capable of cracking Net-NTLMv1 authentication hashes in under 12 hours using consumer-grade hardware. This move aims to highlight the severe, long-known vulnerabilities of the protocol, which has been deprecated for over two decades. Mandiant urges organizations to disable Net-NTLMv1 immediately through Group Policy and monitor for its continued use, emphasizing that credential theft via this method is “trivial.” <a href="https://app.daily.dev/posts/2q2phjvWS" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Cloudflare boosts resilience and acquires Astro</strong>: Cloudflare launched its “Code Orange: Fail Small” initiative to prevent cascading failures after two major global outages in late 2025. The plan includes phased configuration rollouts and automatic rollbacks, as detailed in company statements. Concurrently, Cloudflare acquired The Astro Technology Company, ensuring the Astro framework remains open-source and platform-agnostic, allowing it to focus on framework improvement without monetization pressure. <a href="https://app.daily.dev/feed-by-ids?id=OdNHxfcwg&amp;id=yRAOJZmiR" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>New CSS features and web platform updates</strong>: The Chrome team’s CSS Wrapped 2025 showcases 22 new CSS features, including advancements in Masonry layouts and practical uses for <code>overflow: clip</code>, as highlighted by CSS Weekly. <a href="https://app.daily.dev/posts/dNuWHdQLH" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>React developments and AI coding agents</strong>: Vercel released a React best practices guide for AI coding agents, while discussions emerged on reconstructing React components from production apps using Fiber internals and critical analysis of React 19’s <code>useOptimistic</code> hook, according to React Status. <a href="https://app.daily.dev/posts/RFKAnO4tL" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Go 1.26 RC2 released with security updates</strong>: Go 1.26 Release Candidate 2 is out, bringing security fixes for <code>archive/zip</code> and <code>Request.ParseForm</code>, alongside updates for Go 1.25.6 and Go 1.24.12, as reported by Golang Weekly. <a href="https://app.daily.dev/posts/DFYbtJ1rQ" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI SQL generation in Google BigQuery</strong>: Google is previewing “Comments to SQL” in BigQuery, using Gemini AI to convert natural-language instructions in SQL comments into executable queries, aiming to accelerate data analysis. <a href="https://app.daily.dev/posts/5AcBUpc2i" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>LLM Compressor 0.9.0 enhances model quantization</strong>: LLM Compressor 0.9.0 introduces refactored attention and KV cache quantization supporting FP4, INT8, and FP8, along with Intel’s AutoRound algorithm and experimental MXFP4 support, significantly improving AI model optimization workflows. <a href="https://app.daily.dev/posts/Y9B2mh3yd" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Docker support for Laravel developer tools</strong>: Laravel VS Code Extension and Laravel Boost, both at v1.4.2, now support Docker containerized development environments and include fixes for Pint code formatter issues in DDEV. <a href="https://app.daily.dev/posts/SUaYBzDS2" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Qt Quick and Flutter architectural shifts</strong>: Qt 6.9 and 6.10 enhance Android app integration by allowing embedding of Qt Quick content in Java/Kotlin apps, while Flutter is decoupling Material and Cupertino design systems from its core SDK for better modularity and smaller app sizes. <a href="https://app.daily.dev/feed-by-ids?id=5YFCJbOU9&amp;id=T07WzSMOC" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>AI-generated browser: 3 million lines of code, zero functionality</strong>: Cursor CEO Michael Truell proudly announced that hundreds of GPT-5.2 agents generated over 3 million lines of code in a week to build a web browser. The catch? The code doesn’t compile, has 34 errors and 94 warnings, and there’s no working demo. It seems “meaningful progress” now means generating a mountain of broken code, which is a bold new definition of success, as pointed out by Hacker News. <a href="https://app.daily.dev/feed-by-ids?id=UbZlY7YDE&amp;id=qHCb0ov8q&amp;id=V3KHuMgUU" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI healthcare gold rush: Misinformation and security, but hey, investments!</strong>: OpenAI bought a health startup, Anthropic launched a healthcare AI model, and MergeLabs raised a quarter-billion dollars. All this amidst “major concerns” about AI hallucinating in clinical settings, spreading medical misinformation, and creating security vulnerabilities for patient data. Apparently, the “AI healthcare gold rush” is more about the gold than the actual healthcare, because who needs reliable diagnoses when you have investor enthusiasm? <a href="https://app.daily.dev/posts/YNV6XIM0c" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>The disappearing junior developer: Just become an ‘AI architect,’ easy!</strong>: Junior developer roles are down 46-67% because AI automates entry-level tasks, so aspiring devs should just become “AI architects” and orchestrate AI systems. The solution to job displacement isn’t more entry-level opportunities, but for everyone to instantly become a senior AI expert. Simple, right? Just pivot your entire career to a role that barely existed five minutes ago, and ignore the fact that the AI doing the coding still needs human oversight, as if that’s not a junior-level task. <a href="https://app.daily.dev/posts/dc9HNh5yX" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
