<h2>Must know</h2>
<ul>
<li><strong>Google’s embedding model deprecation causes vendor lock-in</strong>: Google deprecated its <code>text-embedding-004</code> model, rendering thousands of stored vectors unusable for a production RAG-based movie recommender. This highlights a significant vendor lock-in issue, as different embedding models create incompatible vector spaces, making migration difficult. The author had to re-embed their entire dataset and re-validate retrieval quality, emphasizing the need to design for change when persisting embeddings at scale. <a href="https://app.daily.dev/posts/xhDI1zj4i" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Databricks launches Lakebase for AI workloads</strong>: Databricks launched Lakebase, a new PostgreSQL-compatible database specifically designed for AI workloads. It aims to merge operational database capabilities with their existing data lakehouse architecture, offering a serverless OLTP database with features like instant data branching and unified access controls. This move signals Databricks’ push to integrate transactional data directly into their AI platform without the need for ETL. <a href="https://app.daily.dev/posts/56RHAShlB" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Linux kernel 7.0 rc1 released with Rust support</strong>: The Linux kernel 7.0 rc1 has been released, with Linus Torvalds humorously attributing the major version bump to his confusion with large numbers, not a significant feature milestone. This release includes cemented Rust support, faster cache clearing, non-disruptive kernel updates, and updated AMD/Intel silicon support. Torvalds also mused about a future maintainer, referencing the community’s recently developed succession plan. <a href="https://app.daily.dev/posts/cYP8biu57" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>Microsoft’s Project Silica advances glass storage</strong>: Microsoft’s Project Silica made a breakthrough in glass-based data storage, extending it from expensive fused silica to common borosilicate glass, with data longevity estimated at 10,000 years. <a href="https://app.daily.dev/posts/KqLBpGspa" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>New VS Code extension for recent files</strong>: A new VS Code extension, ‘Fresh File Explorer,’ replaces the standard file explorer to display recently modified files, integrating Git history and offering features like a recency heatmap and one-click restore for deleted files. <a href="https://app.daily.dev/posts/pQedWHBRB" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Non-coding tasks hinder developer productivity</strong>: A developer’s personal account highlights that a 10-second code fix can take over 10 hours due to non-coding tasks like fighting sandboxed environments, CUDA conflicts, and build system issues, which current LLM tools don’t address. <a href="https://app.daily.dev/posts/OBIqfWI7x" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Fileless privilege escalation in hybrid AD environments</strong>: A red team case study details a 13-step fileless privilege escalation attack chain in hybrid Active Directory environments, allowing a low-privileged user to gain Domain Admin and Azure AD Global Admin access by exploiting misconfigurations. <a href="https://app.daily.dev/posts/u1TVqzV8C" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Quantum error correction advances with lattice surgery</strong>: Researchers at ETH Zurich made progress in quantum error correction with a technique called lattice surgery, which uses microwave pulses to measure relationships between quantum data blocks, avoiding wave function collapse. <a href="https://app.daily.dev/posts/t1NKbgnH0" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>RwLock underperforms Mutex in Rust read-heavy workloads</strong>: A hands-on investigation found that <code>RwLock</code> performed significantly worse than <code>Mutex</code> for a read-heavy LRU tensor cache in Rust due to cache-line ping-pong, suggesting <code>Mutex</code> might be better for very short critical sections. <a href="https://app.daily.dev/posts/rFMs5918e" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>UMMC hit by ransomware attack</strong>: The University of Mississippi Medical Center suffered a ransomware attack, forcing the shutdown of all 35 clinics, compromising their EHR and IT systems, and leading to canceled elective procedures. <a href="https://app.daily.dev/posts/44GuQSR8k" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>Anthropic’s Claude compiler video: more hype than reality?</strong>: Anthropic released a marketing video showcasing Claude autonomously building a C compiler capable of compiling Linux, SQLite, Lua, and Doom. Critics, however, are calling out the “dishonest framing,” suggesting the actual goal was merely a stress test of Claude’s ability to run autonomously for an extended period, not a straightforward demonstration of compiler-building prowess. So, another day, another AI company trying to make their LLM sound like Skynet. <a href="https://app.daily.dev/posts/vGYk0DbTO" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Meta patents AI to simulate deceased users</strong>: Meta was granted a patent for an AI system designed to simulate a user’s social media activity, including posting and chatting, using an LLM trained on their past data when the user is ‘absent’ from the platform. Because nothing says ‘we care about your mental health’ like having a digital zombie of your dead friend liking your posts. What could possibly go wrong? <a href="https://app.daily.dev/posts/AeW0pHCcd" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Gemini 3.1 Pro: pure disappointment in real-world coding</strong>: A hands-on comparison found Google’s Gemini 3.1 Pro consistently underperformed Claude Opus 4.6 in coding and UI generation tasks, despite strong benchmark scores. Gemini struggled with tool-calling failures, incomplete features, and complex prompts. Apparently, those impressive benchmark numbers don’t translate to actual usefulness. Who would’ve thought? <a href="https://app.daily.dev/posts/bIAFUawfa" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
