<h2>Must know</h2>
<ul>
<li><strong>OpenAI introduces ads to free and Go tiers</strong>: OpenAI is rolling out targeted ads in the free and Go tiers of ChatGPT for US users, a move CEO Sam Altman previously called a “last resort.” This shift, reported by TechCrunch, aims to sustain free access and fund AGI research, following over $11.5 billion in losses in Q3 2025. While OpenAI assures ads won’t influence chatbot answers or sell user data, privacy advocates warn of potential trust issues and problematic incentives. <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 on consumer hardware. This protocol has been deprecated as insecure for over two decades, yet remains prevalent. Mandiant’s release, hosted on Google Cloud, aims to push organizations to finally disable Net-NTLMv1 and implement monitoring, as credential theft via this method is now “trivial.” <a href="https://app.daily.dev/posts/2q2phjvWS" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Cloudflare launches resilience initiative and acquires Astro</strong>: Cloudflare has initiated “Code Orange: Fail Small,” a resilience program designed to prevent cascading failures after two major global outages in late 2025. This involves phased configuration rollouts and automatic rollbacks. 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>AI firms accelerate healthcare expansion</strong>: OpenAI acquired health startup Torch for $100 million, Anthropic launched a Claude AI model for healthcare, and MergeLabs raised $250 million, signaling a rapid acceleration in AI healthcare investment despite concerns about misinformation and security. <a href="https://app.daily.dev/posts/YNV6XIM0c" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Google previews AI SQL generation in BigQuery</strong>: Google is testing a “Comments to SQL” feature in BigQuery, using Gemini AI to convert natural-language instructions in SQL comments into executable queries, aiming to speed up 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 multiple schemes (FP4, INT8, FP8), integrates Intel’s AutoRound, and adds experimental MXFP4 support, significantly improving efficiency for complex AI models. <a href="https://app.daily.dev/posts/Y9B2mh3yd" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Claude Code implements dynamic tool loading for MCP servers</strong>: Claude Code introduced dynamic tool loading for MCP servers, reducing token usage by up to 95% and improving tool selection accuracy by only loading relevant tools on demand, addressing previous context overflow issues. <a href="https://app.daily.dev/feed-by-ids?id=WwZicivI3&amp;id=uq0isMFwF" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AMD EPYC 8004 “Siena” CPUs improve with new Linux stacks</strong>: AMD EPYC 8004 “Siena” processors show significant performance gains with newer Linux distributions and software, particularly Ubuntu 25.10 and GCC 15.2, benefiting cost-optimized server deployments. <a href="https://app.daily.dev/posts/85IZs9OZ7" 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>, with similar updates also released for Go 1.25.6 and Go 1.24.12. <a href="https://app.daily.dev/posts/DFYbtJ1rQ" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AWS expands cloud services to new regions</strong>: Amazon Managed Workflows for Apache Airflow (MWAA) is now available in Asia Pacific (Thailand), and Amazon S3 Storage Lens has expanded to AWS GovCloud (US) Regions, enhancing regional cloud capabilities. <a href="https://app.daily.dev/feed-by-ids?id=9wp5iKTPz&amp;id=tqwyBGRLz" 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’s CEO proudly announced that hundreds of GPT-5.2 agents generated over 3 million lines of code for a web browser in a week, making “meaningful progress.” Turns out, this monumental achievement results in a codebase with 34 errors and 94 warnings that doesn’t even compile. It’s almost like throwing a million monkeys at a keyboard doesn’t magically produce a functional browser, no matter how advanced your AI is. <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>WordPress gets “zero-code” structured data for AI, because SEO wasn’t enough</strong>: ACF PRO 6.8 Beta 2 now offers “zero-code” structured data for WordPress, promising to future-proof sites for “AI-powered content and rich search display demands,” specifically mentioning ChatGPT and Google AI Overviews. Because what every website needs is more automatically generated JSON-LD to feed the AI overlords, ensuring your content is perfectly optimized for whatever hallucination engine is currently trending. No code changes required, just pure, unadulterated <a href="http://Schema.org" target="_blank" rel="noopener nofollow">Schema.org</a> goodness. <a href="https://app.daily.dev/posts/fcLlptMlt" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>The disappearance of the junior developer: blame AI, not the market</strong>: Apparently, junior developer roles have plummeted by 46-67% because AI is automating entry-level coding tasks, forcing aspiring devs to become “AI architects.” So, instead of learning to code, you now need to orchestrate AI systems and manage context windows. It’s a convenient narrative that shifts the blame from a tightening job market and unrealistic expectations to the ever-present scapegoat of AI, ensuring a fresh batch of “AI engineers” for B2B sectors at “premium wages.” <a href="https://app.daily.dev/posts/dc9HNh5yX" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
