<h2>Must know</h2>
<ul>
<li><strong>OpenAI’s aggressive monetization strategy for ChatGPT</strong>: OpenAI has launched ChatGPT Go, an $8/month subscription tier with ads, globally after an initial rollout in India. This move is a direct response to competitive pressure from Google Gemini, which has significantly eroded OpenAI’s market share. The company aims to generate $1 billion in ad revenue in 2026, with projections reaching $25 billion by 2029, by leveraging user conversation data for ad targeting, a strategy that raises significant privacy and ethical concerns, as reported by ‘The A in AGI stands for Ads’ and ‘wtf is going on with ChatGPT?’. <a href="https://app.daily.dev/feed-by-ids?id=ArnlILqap&amp;id=ep7p1hnQy" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>China’s massive satellite internet ambitions and Starlink’s vulnerabilities</strong>: China has filed plans with the ITU to launch two mega-constellations, CTC-1 and CTC-2, totaling nearly 200,000 satellites, dwarfing Starlink’s current plans. This aggressive move aims to dominate orbital radio spectrum and raises national security concerns for the U.S., prompting accelerated FCC approvals for American satellite networks. Meanwhile, Iran successfully disrupted Starlink internet during a national blackout using military-grade GPS jamming, degrading performance by up to 80% and highlighting the system’s vulnerability to state-level electronic warfare, according to reports. <a href="https://app.daily.dev/feed-by-ids?id=9Dqo8zul1&amp;id=VVK9EFJ8w" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI accelerates software development but demands human oversight</strong>: AI-powered code generation is speeding up software development, but human review remains critical for correctness, security, and architectural compliance. The bottleneck has shifted from writing code to reviewing it, requiring engineers to prioritize understanding and responsibility over sheer speed. Concerns about AI-generated code quality, copyright risks, and environmental costs are leading to grassroots projects like OpenSlopware, which tracks LLM-generated code in open-source projects, as discussed on HackerNews and other developer communities. <a href="https://app.daily.dev/feed-by-ids?id=JjrjV7tSv&amp;id=ZeB3UcRxG&amp;id=alpyX7cTa&amp;id=m1Q81Q3xQ" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>Google Antigravity expands beyond coding</strong>: Google’s Antigravity is now competing with Anthropic’s Co-work by adding productivity features like creating Google Slides and organizing files directly from the IDE, leveraging Gemini 3 Flash for simpler tasks. <a href="https://app.daily.dev/posts/HVlI4ECqg" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Wine 11.0 stable release brings major improvements</strong>: Wine 11.0 stable is out with 6,300 code changes, a new NTSYNC kernel module, a complete WoW64 architecture overhaul, and EGL as the default backend for OpenGL, enhancing compatibility and performance for Windows applications on Linux. <a href="https://app.daily.dev/posts/5LDn1X8et" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Wine patches enable direct Adobe Photoshop installation on Linux</strong>: New Wine patches allow direct installation of Adobe Photoshop 2021 and 2025 on Linux by fixing XML parsing issues related to legacy Internet Explorer dependencies, though these patches are not yet in mainline Wine. <a href="https://app.daily.dev/feed-by-ids?id=WW9lhhDTb&amp;id=jcudSc8Dd" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Jakarta EE 12 development and community events update</strong>: The Jakarta EE Platform project is progressing towards Milestone 2 for Jakarta EE 12, with a target date of January 27, 2026, alongside a ramp-up of community conferences and publicly available JakartaOne Livestream 2025 videos. <a href="https://app.daily.dev/posts/YrhkKGKoN" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Linux kernel 6.19-rc6 brings sound fixes</strong>: The latest Linux kernel release candidate, 6.19-rc6, includes crucial sound fixes and workarounds for the ASUS ROG Xbox Ally X gaming handheld and various newer laptop models. <a href="https://app.daily.dev/posts/57EeLYyEN" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Curl removes OpenSSL-QUIC backend</strong>: The curl project is dropping support for the OpenSSL-QUIC backend in version 8.19.0 due to its inadequate API, poor performance (3x slower), and excessive memory usage (20x more) compared to alternatives, consolidating HTTP/3 support to nghttp2+nghttp3 and quiche. <a href="https://app.daily.dev/posts/As2f3e0AW" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Pharmaverse reduces CI/CD time with Docker containers</strong>: The pharmaverse blog cut its CI/CD publishing time from 17 to 5 minutes by switching from on-the-fly R package installation to using a pre-built Docker container image with over 40 pre-installed packages. <a href="https://app.daily.dev/posts/ZDzd1Eyhy" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>AI will get malloc wrong forever, apparently</strong>: Some folks are claiming that AI code generation tools will ‘forever’ get memory management wrong, specifically with <code>malloc</code>, because LLMs statistically mirror flawed patterns in popular codebases. The argument is that since <code>malloc</code> doesn’t zero out memory, AI-generated code often returns allocated buffer sizes instead of actual data, creating security vulnerabilities. So, if you’re relying on AI for low-level C, prepare for some ‘fun’ debugging sessions with uninitialized memory leaks, because apparently, statistical models can’t grasp basic memory safety. <a href="https://app.daily.dev/posts/x9XQt7sEo" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI-powered DevOps: Just add YAML generation and troubleshooting</strong>: A new 4-month live course promises to teach ‘AI-Powered DevOps with AWS,’ which apparently means adding AI for YAML generation and troubleshooting to your standard Linux, Docker, Kubernetes, and Jenkins stack. Because nothing says ‘AI-powered’ like automating the most tedious parts of configuration files and having a bot tell you why your pipeline broke. It’s almost as if the ‘AI’ part is just a sprinkle on top of existing DevOps practices to make it sound cutting-edge. <a href="https://app.daily.dev/posts/ZYeWLps6B" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>RFC-driven AI coding: The bureaucracy of the future</strong>: Someone created a CLI tool called <code>govctl</code> to enforce ‘RFC-driven AI coding,’ demanding specifications before implementation and phase gates for AI-generated code. Because what AI-assisted coding really needed was more bureaucracy and process, not just, you know, good code. It’s designed for teams who want ‘structured, spec-first development rather than rapid prototyping,’ which sounds like a fancy way of saying ‘we don’t trust the AI, but we’re still going to use it, just with more paperwork.’ <a href="https://app.daily.dev/posts/oBJvqLGNT" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
