<h2>Must know</h2>
<ul>
<li><strong>Meta shifts from metaverse to AI, cuts Reality Labs staff</strong>: Meta is closing three VR studios and laying off 10% of its Reality Labs workforce, approximately 1,500 employees, after the division lost over $70 billion since 2021. This strategic pivot, reported by multiple sources, reallocates resources from VR and metaverse initiatives towards AI-powered glasses and wearable technology. The move signals a significant reorientation for Meta, which had previously centered its corporate rebranding on the metaverse. <a href="https://app.daily.dev/posts/7ystcAQPc" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Cloudflare fined by Italy over DNS refusal, threatens withdrawal</strong>: Cloudflare has been fined €14.2 million by Italy’s AGCOM for refusing to block pirate sites on its 1.1.1.1 DNS resolver, as mandated by Italy’s Piracy Shield law. Cloudflare argues compliance is technically infeasible without impacting all users and threatens to withdraw its servers from Italy. CEO Matthew Prince is seeking U.S. government support and has notified the International Olympic Committee about potential cybersecurity risks for the 2026 Winter Olympics in Italy, escalating the dispute. <a href="https://app.daily.dev/posts/Q0IbcQd4P" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Microsoft AI data center expansion faces backlash</strong>: Microsoft’s ‘Community-First AI Infrastructure’ initiative, promising to cover power costs and minimize water usage for its U.S. data centers, is a direct response to widespread public backlash. Activist groups in 24 states have protested the environmental and social impacts of these expansions, leading Microsoft to abandon projects in Wisconsin and face ongoing resistance in Michigan and Ohio. The IEA projects global data center electricity demand to more than double by 2030, intensifying scrutiny on AI’s energy footprint. <a href="https://app.daily.dev/posts/jqtBl3Jnb" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>Anthropic commits $1.5 million to Python Software Foundation</strong>: Anthropic has pledged $1.5 million over two years to the Python Software Foundation (PSF) to enhance Python ecosystem security, focusing on automated malware detection for PyPI packages and supporting CPython core development. <a href="https://app.daily.dev/posts/yCHi0TB7q" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>New Linux malware targets cloud and container infrastructure</strong>: Check Point Research identified VoidLink, a new Linux malware framework written in Zig, designed to infiltrate cloud and container environments (AWS, Azure, GCP, Docker, Kubernetes) using advanced rootkit techniques and a modular plugin system for credential harvesting and lateral movement. <a href="https://app.daily.dev/posts/uJaIfhgHH" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Go-based LLM gateway Bifrost claims 40x speedup over LiteLLM</strong>: Bifrost, an open-source LLM gateway written in Go, offers unified access to over 15 AI providers via an OpenAI-compatible API, claiming 40x lower gateway overhead and 68% lower memory usage than LiteLLM, with features like semantic caching and adaptive load balancing. <a href="https://app.daily.dev/posts/P8RYoHcNL" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Rust infrastructure team Q4 recap and Q1 plan</strong>: The Rust Infrastructure Team successfully migrated to the new Bors bot for CI, moved <a href="http://crates.io" target="_blank" rel="noopener nofollow">crates.io</a> index and <a href="http://docs.rs" target="_blank" rel="noopener nofollow">docs.rs</a> to Fastly CDN, and expanded Infrastructure as Code coverage, with Q1 2026 plans including <a href="http://docs.rs" target="_blank" rel="noopener nofollow">docs.rs</a> modernization and external hardware CI policy. <a href="https://app.daily.dev/posts/m2OVQmP6S" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>LLVM/Clang 22 adds Intel Nova Lake and Arm C1 support</strong>: LLVM/Clang 22 feature development concluded, adding support for Intel Nova Lake and Wildcat Lake processors, Arm C1 series, and Ampere1C CPUs, alongside C2y language features and Zen 4 optimizations, with a 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>Node.js releases address security vulnerabilities</strong>: Node.js 20.20.0 (LTS) and 25.3.0 (Current) were released, addressing multiple CVEs by disabling futimes with the permission model, adding default TLSSocket error handlers, and tightening symlink API permissions. <a href="https://app.daily.dev/feed-by-ids?id=DwsoMgMKH&amp;id=Zr0Dv9dwZ" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>WordPress 7.0 to drop PHP 7.2/7.3 support</strong>: The WordPress core development team plans to discontinue support for PHP 7.2 and 7.3 in version 7.0, with a meeting scheduled to finalize details and call for volunteers for the release. <a href="https://app.daily.dev/posts/xho5u0zys" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>AI agents are unreliable and insecure by design, says Signal</strong>: Signal’s leadership is out here warning that agentic AI systems, like Microsoft’s Recall, are ‘unreliable and insecure by design,’ citing prompt injection attacks and vulnerable user activity storage. They argue these AI agents are probabilistic, not deterministic, meaning failures compound and accuracy plummets for multi-step tasks. So, while everyone else is hyping AI’s boundless potential, Signal is basically saying, ‘Yeah, but it’s still mostly a buggy mess that can’t be trusted.’ <a href="https://app.daily.dev/feed-by-ids?id=TJt6uaOsV&amp;id=XnmeLgefE&amp;id=rjeAcjlYe" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Anthropic’s token crackdown: ‘No way’ to use our AI with your tools</strong>: Anthropic, in a move that screams ‘we need more money,’ is now strictly enforcing token restrictions, preventing users from using Claude Code with third-party tools like Cursor and Open Code. Apparently, this was always in their terms of service, but they just decided to start caring now that developers have built entire workflows around it. Developers are, predictably, ‘controversial’ and ‘backlash’-ing, because nothing says ‘developer-friendly’ like pulling the rug out from under them. <a href="https://app.daily.dev/feed-by-ids?id=DEqFrm707&amp;id=jcpFkjQkx" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI will revolutionize ecology, but who needs actual ecologists?</strong>: Ecologists are apparently ditching their boots for AI, machine learning, and automated sensors to study biodiversity. While this allows for ‘broader and more frequent data collection,’ some scientists are ‘concerned’ about losing hands-on understanding. It’s almost as if replacing human expertise with algorithms might have some downsides, but hey, at least we’ll have more data, right? Who needs messy fieldwork when you have a drone and a neural network? <a href="https://app.daily.dev/feed-by-ids?id=YsLIINgTC&amp;id=dHxdm7wPU" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
