<h2>Must know</h2>
<ul>
<li><strong>Meta and Google double down on AI infrastructure and models</strong>: Meta is launching ‘Meta Compute,’ a massive infrastructure scale-up aiming for tens of gigawatts of power capacity to support future AI workloads, led by Santosh Janardhan, Daniel Gross, and Dina Powell McCormick. This puts them in direct competition with Microsoft and Google for generative AI cloud dominance. Meanwhile, Google released Gemini 3, a multimodal AI platform with a 1M+ token context window, processing text, images, video, audio, and PDFs in a unified model, promising ‘programming gold medal performance’ and deploying across Google’s application suite, as reported by their official communications. <a href="https://app.daily.dev/feed-by-ids?id=2S31ZjKRT&amp;id=NEGZugNBg&amp;id=kwRMqebwB&amp;id=qjYQj2ENJ" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Anthropic’s Claude Code token restrictions spark developer backlash</strong>: Anthropic has begun strictly enforcing its terms of service, preventing the use of Claude Code subscription tokens with third-party tools like Cursor and Open Code. This forces developers to use Claude Code exclusively or pay significantly higher API prices, sparking controversy among those who built workflows around these integrations. The move is seen as an attempt to lock developers into Anthropic’s ecosystem, potentially driven by unsustainable infrastructure costs, according to technical reporting. <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>Microsoft’s AI data center expansion faces community and energy scrutiny</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 in 24 states. Despite pledges to avoid local tax breaks and prevent residential electricity bill increases, community opposition has already led Microsoft to abandon projects in Wisconsin, with Michigan and Ohio facing ongoing resistance. This highlights growing consumer concerns about AI infrastructure’s impact on energy rates and water supplies, as reported by various sources. <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>Open-weight AI models rival frontier models at lower cost</strong>: New open-weight AI models, GLM 4.7 and MiniMax M2.1, demonstrated the ability to autonomously build complex CLI task runners, with GLM 4.7 completing the task in 10 minutes and MiniMax M2.1 at half the cost, showcasing their competitive capabilities against closed models. <a href="https://app.daily.dev/feed-by-ids?id=p3khbP8jU&amp;id=sDf9ENrg4" 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 like AWS, Azure, GCP, Docker, and 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>Salesforce rebrands Slackbot as an AI agent</strong>: Salesforce has relaunched Slackbot as an AI agent for Business+ and Enterprise+ customers, leveraging context from Slack and third-party tools to answer questions, organize work, create content, and schedule meetings via natural language prompts. <a href="https://app.daily.dev/posts/VpXLIsjSp" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI agent builders get go-to-market infrastructure</strong>: Aden is a new SDK providing white-labeled portals, trust centers with audit trails, and budget caps for AI agent builders, aiming to help them meet enterprise security requirements and accelerate deals. <a href="https://app.daily.dev/posts/ILaKe7jt3" 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 with security updates addressing multiple CVEs, including disabled futimes when the permission model is enabled, tightened symlink API permissions, and improved TLSSocket error handling. <a href="https://app.daily.dev/feed-by-ids?id=DwsoMgMKH&amp;id=Zr0Dv9dwZ&amp;id=Wy4br4Zyw" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>WordPress 7.0 to drop PHP 7.2 and 7.3 support</strong>: The WordPress core development team is planning to discontinue support for PHP 7.2 and 7.3 in the upcoming WordPress 7.0 release, with a call for volunteers to participate in the release process. <a href="https://app.daily.dev/posts/xho5u0zys" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Bun v1.3.6 introduces tarball handling and performance boosts</strong>: Bun v1.3.6 launched with Bun.Archive for tarball creation and extraction, a JSONC parser, and significant performance improvements, including 3.5x faster Response.json() and 15% faster async/await operations. <a href="https://app.daily.dev/feed-by-ids?id=Wx4br4Zyw&amp;id=Yy5NFIIOM" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>NVIDIA CEO thinks AI criticism is a ‘doomer narrative’</strong>: Jensen Huang, NVIDIA’s CEO, apparently believes ‘well-respected people’ criticizing AI are just spreading a ‘doomer narrative’ that stifles startups. He’s questioning the intentions of PhDs and CEOs who warn governments about dystopian AI scenarios. Funny how the guy whose company is making billions off the AI boom thinks everyone else should just shut up and enjoy the ride, as reported by 80 Level. Maybe he’s just worried about regulations cutting into his military contracts? <a href="https://app.daily.dev/posts/1vSR1HgdF" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI agents are unreliable and insecure, says Signal</strong>: Signal’s president, Meredith Whittaker, and VP are out here saying agentic AI systems, like Microsoft’s Recall, are ‘unreliable and insecure by design.’ They claim these probabilistic systems fail at each step, making multi-step tasks a joke, and are vulnerable to prompt injection attacks. So, while Satya Nadella is busy promoting ‘productivity gains,’ Signal is calling for a halt to this ‘reckless deployment’ and demanding transparency. Who knew that AI agents might not be the infallible overlords we were promised? <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>AI coding costs are eating into developer salaries</strong>: One developer, after blowing $100 on Claude Opus tokens in a single day, is now claiming we’ve all been ‘played for fools.’ Apparently, these high token consumption costs are effectively reducing our income by about 10%. So much for AI making our lives easier and cheaper, right? Maybe ‘vibe coding’ isn’t as free-spirited as it sounds when Anthropic is quietly siphoning off your paycheck, as shared by ThePrimeTime. <a href="https://app.daily.dev/posts/UkrFq25pN" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
