<h2>Must know</h2>
<ul>
<li><strong>AI coding assistants: more planning, more bugs, more cost</strong>: AI coding assistants, while perceived as productivity boosters, actually demand significantly more upfront architectural planning and specification work. A 2025 study revealed that teams using AI assistants took 19% longer to complete tasks, with code review times increasing by 91% and 45% of AI-generated code containing security vulnerabilities. This suggests that without robust specifications and human oversight, AI can introduce an ‘architecture tax’ that negates perceived gains, as reported by multiple sources. <a href="https://app.daily.dev/feed-by-ids?id=pYjmmMIkn&amp;id=MVxk3C8r1&amp;id=2tsKdG55u" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>EU investigates Grok AI for harmful image generation</strong>: The European Commission has launched an investigation into Elon Musk’s Grok AI over allegations it generated sexually explicit images, including those of minors. This action falls under the EU’s Digital Services Act (DSA), which mandates strict online protections for children. X, Grok’s owner, has since disabled the image generation feature for non-paying users, but faces potential fines up to $174 million if violations are confirmed, according to reports from multiple sources. <a href="https://app.daily.dev/feed-by-ids?id=tpCDWbga3&amp;id=KoLpeTHv0" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>TSMC’s flat CapEx creates AI chip bottleneck for hyperscalers</strong>: Despite a surge in demand for AI chips following ChatGPT’s launch, TSMC maintained flat capital expenditures from 2021 to 2024, leading to significant supply bottlenecks. This imbalance is projected to cost hyperscalers billions in lost AI revenue, with shortages expected to persist through at least 2028. Analysts suggest hyperscalers should diversify their supply chains to mitigate ongoing risks, as reported by industry experts. <a href="https://app.daily.dev/posts/pPxEIulIE" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>Nvidia invests $2B in CoreWeave for AI compute expansion</strong>: Nvidia is investing $2 billion in CoreWeave to accelerate the construction of over 5 gigawatts of AI computing infrastructure by 2030, integrating Nvidia’s Rubin chip architecture and other technologies into CoreWeave’s platform. <a href="https://app.daily.dev/posts/suyuzcb2U" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>OpenAI upgrades ChatGPT code execution with Bash and package support</strong>: OpenAI has enhanced ChatGPT’s code execution feature, allowing users to run Bash commands and install packages via pip and npm within its sandboxed environment, significantly expanding its utility for developers. <a href="https://app.daily.dev/posts/7V1v7MpYn" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Docker Sandboxes isolates AI coding agents with microVMs</strong>: Docker Sandboxes now uses microVM-based isolation for AI coding agents like Claude Code and Gemini CLI, allowing them to install packages and modify files without affecting the host system, currently available on macOS with Windows and Linux support planned. <a href="https://app.daily.dev/posts/6YdOupFK7" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Google launches FunctionGemma for mobile/edge AI function calling</strong>: Google released FunctionGemma, a 270M parameter model optimized for function calling on mobile and edge devices, enabling AI agents to translate natural language into structured API calls for on-device task automation. <a href="https://app.daily.dev/posts/DFyxBQQFb" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Nvidia releases Earth-2 AI weather models as open source</strong>: Nvidia has open-sourced three Earth-2 AI weather forecasting models—Nowcasting, Medium Range, and Global Data Assimilation—aiming to democratize access to advanced forecasting and reduce reliance on government supercomputers. <a href="https://app.daily.dev/posts/AZxwXEsfz" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Microsoft launches Maia 200 AI inference accelerator</strong>: Microsoft unveiled the Maia 200, a custom AI inference accelerator built on TSMC’s 3nm process, featuring FP8 and FP4 tensor cores, 216GB HBM3e memory, and claiming 30% better performance per dollar for tasks like GPT-5.2 models and Microsoft 365 Copilot. <a href="https://app.daily.dev/posts/HBgJElvHn" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Claude adds integrated interactive tools to conversations</strong>: Claude now supports integrated interactive tools within its conversations, allowing users to access applications like Asana, Slack, Figma, and Amplitude directly from its interface, powered by the open Model Context Protocol (MCP). <a href="https://app.daily.dev/posts/ddUJyBQv7" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>AI-generated software tests: the illusion of quality</strong>: Apparently, LLMs are now churning out software tests, but don’t get too excited. According to some ‘experts,’ these tests are just ‘ceremony’ and give a ‘false sense of security’ because they’re not validated by actually failing. So, we’re automating the creation of useless tests. Progress! <a href="https://app.daily.dev/posts/wLM9iDorf" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI coding assistants: privatizing knowledge, one prompt at a time</strong>: The latest hot take is that LLM-powered coding assistants are quietly siphoning off all our precious open-source knowledge into private, paywalled silos. Soon, all the good coding wisdom will be locked behind a subscription, and Stack Overflow will be a ghost town. Because nothing says ‘innovation’ like making knowledge less accessible. <a href="https://app.daily.dev/posts/p61M36Xco" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Cursor AI’s browser: built ‘from scratch’ by AI, barely works</strong>: Cursor’s CEO proudly announced their AI agents built a web browser ‘from scratch’ in a week, generating millions of lines of code. Independent developers, however, found the browser barely compiles, relies heavily on existing projects like Servo, and performs ‘abysmally.’ It seems ‘from scratch’ now means ‘cobbled together by AI with a generous dose of marketing fluff.’ <a href="https://app.daily.dev/posts/DovPZUpOP" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
