<h2>Must know</h2>
<ul>
<li><strong>Kubernetes RCE flaw identified, patch incoming</strong>: Edera Advisory found a critical design flaw in Kubernetes that allows remote code execution (RCE) in any container on a node via <code>nodes/proxy GET</code> permissions. This is a big deal because these permissions are commonly used by monitoring tools. While Kubernetes 1.36, expected in April 2025, will remove the problematic permission, many production clusters won’t update that fast. Organizations should scan for RCE attempts and tighten authorization controls, as reported by Edera Advisory. <a href="https://app.daily.dev/posts/phMyeDnVg" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Musk’s companies eye massive mergers</strong>: Elon Musk’s SpaceX, xAI, and potentially Tesla are reportedly in early talks to merge, aiming for a trillion-dollar super-company. The idea is to combine rockets, satellites, AI, and EVs, with a key driver being space-based AI data centers to cut training costs. This could also strengthen Pentagon contracts. SpaceX’s IPO is planned for 2026, and xAI recently acquired X, making this a significant consolidation play. <a href="https://app.daily.dev/posts/wMYF638eZ" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Microsoft’s hybrid AI chip strategy</strong>: Microsoft has deployed its Maia 200 AI inference chip in data centers, claiming superior performance over competitors. However, CEO Satya Nadella confirmed they’ll keep buying AI chips from Nvidia and AMD, indicating a diversified supply chain. The Maia 200 will initially support Microsoft’s Superintelligence team and OpenAI’s models on Azure, highlighting the ongoing challenge of securing advanced AI hardware. <a href="https://app.daily.dev/posts/kN85HgrLU" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>Container security still a mess</strong>: A BellSoft survey found 23% of developers experienced container security incidents last year, with human error causing 62% of mistakes, and nearly half prefer pre-hardened images over managing security themselves. <a href="https://app.daily.dev/posts/I55IVTrKU" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Google Cloud brings conversational analytics to BigQuery</strong>: Google Cloud launched Conversational Analytics in BigQuery (preview), an AI-powered agent in BigQuery Studio that uses Gemini models to generate, execute, and visualize SQL queries from natural language, eliminating the need for users to write SQL. <a href="https://app.daily.dev/posts/2YKMeABdR" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Meta’s AI infrastructure spending skyrockets</strong>: Meta plans to invest $115-135 billion in AI infrastructure in 2026, nearly doubling its 2025 investment, to support its Superintelligence Labs and integrate large language models across its platforms. <a href="https://app.daily.dev/posts/JHGPGk0MW" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Streaming SQL for microservices</strong>: Streaming SQL offers an alternative to traditional microservices by processing unbounded event streams with continuous queries, integrating AI/ML models, and offloading complexity like repartitioning and failure recovery. <a href="https://app.daily.dev/posts/N2LwngQDr" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>New Go linter prevents transaction bugs</strong>: A developer built a custom Go linter using <code>go/analysis</code> to detect when code inside transaction callbacks incorrectly uses outer repository instances, preventing database operation leaks outside transaction boundaries. <a href="https://app.daily.dev/posts/Lx23O016G" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Rust OS for IoT devices</strong>: Ariel OS is a new Rust-based library operating system designed for secure, memory-safe, low-power IoT devices on 32-bit microcontrollers, integrating existing Embedded Rust ecosystem projects. <a href="https://app.daily.dev/posts/eCviEdJvc" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>formae IaC platform adds multi-cloud support</strong>: Platform Engineering Labs announced beta multi-cloud support for its open-source IaC platform, formae, now including Google Cloud, Azure, Oracle Cloud, and OVHcloud, along with a Plugin SDK for custom extensions. <a href="https://app.daily.dev/posts/bqzmDPKsL" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>Google DeepMind’s ‘Project Genie’ is here to save gaming (or not)</strong>: Google DeepMind launched ‘Project Genie,’ an AI world generator that creates ‘infinite, interactive environments.’ Google says it’s a step toward AGI and a game prototyping tool, but 52% of game developers now think AI negatively impacts their field, up from 30% last year. So, while Google’s busy generating whimsical claymation worlds, the actual humans making games are getting laid off. Progress! <a href="https://app.daily.dev/posts/iit38VC43" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Meta Reality Labs: Peak losses, new pivot, same old story</strong>: Meta’s Reality Labs lost another $19.2 billion last year, bringing its total losses to a cool $83.6 billion since 2020. Mark Zuckerberg, ever the optimist, says this is ‘peak losses’ and they’re pivoting to glasses and wearables with AI. Because clearly, throwing more money at a problem and rebranding it with ‘AI’ always works, right? Just ask the 10% of Reality Labs staff who were recently cut. <a href="https://app.daily.dev/posts/ALCffHVkc" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Developers still over-architecting for imaginary scale</strong>: Apparently, some developers are still designing for ‘extreme scalability’ for hundreds of users, wasting resources on problems that don’t exist yet. As one source eloquently put it, this is ‘on the expense of actual problems.’ Maybe focus on building something that works for <em>today’s</em> users before optimizing for the hypothetical millions who might never show up. <a href="https://app.daily.dev/posts/KScX9il0q" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
