<h2>Must know</h2>
<ul>
<li><strong>Mandiant report reveals 22 second ransomware handoff</strong>: Google’s Mandiant M-Trends 2026 report highlights a terrifying collapse in the attack lifecycle, where the window between initial access and ransomware deployment has dropped from 8 hours to just 22 seconds. Voice phishing (vishing) has surged to the second most common entry point, frequently used to trick IT help desks into registering attacker-controlled MFA devices. Defenders are urged to move toward behavioral anomaly detection as the mean time to exploit vulnerabilities is now negative, meaning exploitation often precedes patch releases. <a href="https://app.daily.dev/feed-by-ids?id=YEw08MVRF&amp;id=6lxfwLMSt" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>TypeScript 6.0 bridges the gap to a Go native future</strong>: The release of TypeScript 6.0 marks the final JavaScript-based version of the compiler before the transition to a high-performance Go-native engine in version 7.0. This release introduces breaking changes to tsconfig defaults, including strict mode and ES2025 targets, while the new ‘empty by default’ types setting reportedly cuts build times by 50%. To ease the migration, a new stable type ordering flag has been added to ensure consistency during the jump to the native compiler. <a href="https://app.daily.dev/posts/kynZFJI6r" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Xbox One finally falls to voltage glitching exploit</strong>: Over a decade after its release, the Xbox One’s hardware security has been fully compromised using a technique called ‘Bliss’ that targets the CPU voltage rail. By delivering two precisely timed voltage collapses, researcher Gaasedelen bypassed ARM Cortex memory protections and hijacked a boot ROM Memcpy operation. Because the exploit targets silicon-level code, it is unpatchable and allows for unsigned code execution at the hypervisor and security processor levels, as reported by Schneier on Security. <a href="https://app.daily.dev/posts/59weZvOHj" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>AMD proposes pghot kernel subsystem for tiered memory</strong>: A new Linux kernel subsystem aims to unify hot-page tracking and promotion logic to optimize performance on systems with CXL and multiple memory tiers. <a href="https://app.daily.dev/posts/9z2uCLEQw" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Intel FRED enabled by default in Linux 7.1</strong>: Flexible Return and Event Delivery is now the default for Panther Lake chips, offering lower latency for privilege-level transitions and I/O tasks. <a href="https://app.daily.dev/feed-by-ids?id=G3fsiNuau&amp;id=5l3MiWfx9" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Netflix Graph Abstraction manages 650TB at millisecond latency</strong>: The new internal platform uses a gRPC API inspired by Gremlin to handle massive social and service topology graphs with single-digit millisecond response times. <a href="https://app.daily.dev/posts/esBIPovP0" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Canonical joins Rust Foundation as Gold Member</strong>: The move follows Ubuntu’s aggressive replacement of core GNU components like sudo and coreutils with Rust-based implementations. <a href="https://app.daily.dev/posts/udyWvo5kF" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Unikraft boots one million VMs in ten milliseconds</strong>: Using Firecracker and compressed snapshots, this demonstration shows VM-grade isolation with container-like density for scaling AI agent sandboxes. <a href="https://app.daily.dev/posts/ZIuVqXL43" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Pearl Abyss apologizes for undisclosed AI assets in Crimson Desert</strong>: The studio admitted that 2D visual props generated by AI during early production were ‘unintentionally’ left in the final release version. <a href="https://app.daily.dev/posts/iQy5B7Sjk" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>RE# engine claims O(n) complexity for all regex matches</strong>: A new two-pass regex engine solves the O(n²) complexity issue found in RE2 and Rust’s regex crate while preserving POSIX semantics. <a href="https://app.daily.dev/posts/d2ORtwg2D" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>The 98 percent accuracy dark web hallucination</strong>: Google is claiming its new Gemini-powered threat intelligence service identifies dark web threats with 98% accuracy, a figure that smells like pure marketing fiction given the noisy, slang-heavy nature of underground forums. They contrast this with the 90% false positive rate of ‘traditional’ tools, conveniently ignoring that an AI agent autonomously building an ‘organizational profile’ is just a fancy way of saying it’s guessing based on business context. It’s another ‘AI-native’ security wrapper that promises to replace human analysts with a black box that likely just filters for keywords it was told to ignore. <a href="https://app.daily.dev/posts/40Ri2RlBf" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Roblox claims brand tax isn’t about the money</strong>: Roblox VP Stephanie Latham is trying to convince creators that charging a new fee for brand integrations ‘is not about revenue at all,’ but rather about ‘safety and compliance.’ It’s a classic corporate pivot: taking a 30% cut of in-app purchases wasn’t enough, so now they’re coming for the independent ad deals creators spent years building. Framing a blatant revenue grab as a ‘standardized ad measurement’ service is a masterclass in gaslighting your own developer ecosystem. <a href="https://app.daily.dev/posts/T6pMWbYIp" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Vibe coding leads to 17 percent dumber developers</strong>: An Anthropic study found that while developers using AI agents finished tasks faster, they scored 17% lower on conceptual understanding of the code they ‘wrote.’ This confirms what every senior engineer feared: we’re raising a generation of ‘vibe coders’ who can prompt a solution but can’t explain why it works or how it will fail in production. We’re effectively trading long-term architectural integrity for the short-term dopamine hit of a passing test suite. <a href="https://app.daily.dev/feed-by-ids?id=EQi7fPMBC&amp;id=seweaAUF7" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
