<h2>Must know</h2>
<ul>
<li><strong>Stripe considers acquiring PayPal</strong>: Stripe, now valued at $159 billion after a recent tender offer, is reportedly eyeing PayPal for acquisition. This comes as PayPal, valued at over $40 billion, faces slowing growth and increased competition from tech giants like Apple and Google. While talks are in early stages, the news from Bloomberg has already caused PayPal’s shares to jump. <a href="https://app.daily.dev/feed-by-ids?id=gB7UDBJXo&amp;id=2WYO6dKLA&amp;id=HqWjbO0Bb" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Major npm supply chain attack targets AI coding tools</strong>: Socket researchers uncovered ‘SANDWORM_MODE,’ an active npm supply chain attack using 19 typosquatted packages to impersonate developer utilities and AI coding tools like Claude Code. This malware harvests credentials, injects malicious code into GitHub Actions and CI pipelines, and deploys a malicious MCP server for prompt-injection attacks, silently exfiltrating SSH keys and cloud credentials. npm has responded by tightening registry controls. <a href="https://app.daily.dev/feed-by-ids?id=q6c2RLQKH&amp;id=3fX3MdYNS" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Cybersecurity landscape worsens with AI-enabled attacks</strong>: CrowdStrike’s latest report indicates a dramatic escalation in cyber threats, with average breakout times dropping to 29 minutes and AI-enabled attacks surging by 89%. Threat actors are increasingly using Generative AI across the attack kill chain, targeting AI systems themselves, and cloud intrusions are up 37%. This shift emphasizes stealthy, malware-free attacks, highlighting a critical need for proactive defense strategies. <a href="https://app.daily.dev/posts/WsqeDq8OW" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>Meta secures major AMD chip deal for AI expansion</strong>: Meta signed a multiyear deal with AMD for up to $100 billion in chips, including MI540 GPUs, to add 6GW of data center compute capacity and diversify its AI infrastructure beyond Nvidia. <a href="https://app.daily.dev/posts/vYZNsZv8f" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>OpenAI revises compute target, projects strong revenue and user growth</strong>: OpenAI has lowered its compute spending target to $600 billion by 2030, while projecting over $280 billion in revenue and finalizing a new funding round exceeding $100 billion at a $730 billion valuation. <a href="https://app.daily.dev/posts/JOWJuLHZs" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Apple researchers introduce on-device AI model ferret-UI lite</strong>: Apple researchers developed Ferret-UI Lite, a compact 3B-parameter multimodal AI model designed for efficient, on-device understanding and interaction with user interfaces, prioritizing low latency, privacy, and offline capability. <a href="https://app.daily.dev/posts/qj9GoBvQx" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>New LLM inference technique triples speed</strong>: Researchers developed a multi-token prediction technique that triples LLM inference speed without auxiliary draft models, achieving 3x+ speedup with under 3% accuracy loss on reasoning-heavy tasks by fine-tuning models with a confidence-adaptive decoding strategy. <a href="https://app.daily.dev/posts/hxaAjvILy" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI agent infrastructure, security, and observability developments</strong>: New platforms like Agentuity’s Gravity Network and Tinybird’s Branches are emerging to support AI agent workloads with isolated environments and robust data management, while OpenLLMetry extends OpenTelemetry for AI agent observability. <a href="https://app.daily.dev/feed-by-ids?id=0HG7ge0iD&amp;id=29JQKm5zh&amp;id=4Ak1fvjod&amp;id=EUE0Dit8s&amp;id=F9Y7SyAcI&amp;id=OBZ9FuH7A&amp;id=fWsa69inU&amp;id=HYKgEeyqr&amp;id=ILJ5XhRHr&amp;id=T9FzLQflQ" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Node.js 25.7.0 released with HTTP/2 and SEA improvements</strong>: Node.js 25.7.0 introduces HTTP/2 <code>http1Options</code>, ESM support for Single Executable Applications, a SQLite module release candidate, and various fixes for HTTP keep-alive, FileHandle leaks, and zstd memory tracking. <a href="https://app.daily.dev/posts/K4geCuziz" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>DMARCbis update and mailbox provider requirements</strong>: The DMARCbis update (draft-ietf-dmarc-dmarcbis-41) is evolving the DMARC email authentication standard with clearer reporting expectations and standardized DNS tree walks, as major mailbox providers tighten sender requirements for DMARC, SPF, and DKIM. <a href="https://app.daily.dev/posts/yEFoBF8zN" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>Cursor launches autonomous cloud AI agents for software development</strong>: Cursor is now touting ‘autonomous cloud AI agents’ that can supposedly do your coding, produce ‘merge-ready Pull Requests,’ and even send video demos. Because nothing says ‘trustworthy code’ like an AI that claims to be ‘smooooth’ and replaces traditional code reviews with a video. What could possibly go wrong when we delegate all the actual thinking to a machine? <a href="https://app.daily.dev/posts/BbESdM4Ct" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI agents impact on ‘10x developer’ concept</strong>: Justin Searls, a named individual, confidently declared that AI agents have rendered the ‘10x developer’ debate moot, claiming they equalize or amplify developer output. Conveniently, he offers zero evidence or specifics on how this magical equalization occurs, just a self-congratulatory pat on the back for his ‘nailed’ prediction. Sounds like another expert who’s found a way to make his past opinions sound prophetic. <a href="https://app.daily.dev/posts/dMpIU74OH" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AWS AI tool outage blame controversy</strong>: AWS had an outage in China, and their official story is that an engineer with ‘broader permissions than expected’ accidentally triggered a CloudFormation teardown using Kiro, their AI coding tool. AWS calls Kiro’s involvement a ‘coincidence,’ while simultaneously laying off thousands of human workers. So, the AI is just an innocent bystander, and the solution is more human oversight for AI-generated changes? Classic blame-shifting, protecting the shiny new toy over the actual humans. <a href="https://app.daily.dev/posts/qZoHoApRh" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
