ClickHouse

Frontend Masters

Hacker News

InfoQ

InfoWorld

<h2>Must know</h2>
<ul>
<li><strong>TSMC sees endless AI chip demand, plans massive expansion</strong>: TSMC reported record Q4 earnings for 2025 and forecasts 30% revenue growth for 2026, driven by what its CEO calls “endless” demand for AI chips. The company plans $52-56 billion in capital expenditures for 2026, with most going to advanced 7nm and smaller process nodes, including a rapid ramp-up of 2nm production. This expansion, detailed in TSMC’s official financial reporting, includes 30% of 2nm production in the USA, signaling a strategic push to meet the insatiable appetite for AI infrastructure. <a href="https://app.daily.dev/posts/aTc9ls9yL" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Critical WordPress plugin vulnerability grants instant admin access</strong>: A critical privilege escalation vulnerability (CVE-2026-23550) in the Modular DS WordPress plugin, affecting versions 2.5.1 and below, allows unauthenticated attackers to gain instant admin access. The flaw, with a CVSS score of 10.0, stems from faulty API routing logic. Active exploitation has been detected, and with over 40,000 sites estimated to be vulnerable, immediate updates to version 2.5.2 are advised, along with auditing for unauthorized admin accounts. <a href="https://app.daily.dev/posts/D6wKV3WK1" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Palo Alto Networks patches critical denial-of-service flaw</strong>: Palo Alto Networks has released patches for CVE-2026-0227, a high-severity denial-of-service vulnerability in PAN-OS firewalls with GlobalProtect enabled. This flaw allows unauthenticated attackers to force firewalls into maintenance mode, potentially causing network outages. The existence of public proof-of-concept code increases the urgency for immediate patching, especially given a similar zero-day exploit in 2024, as reported directly by Palo Alto Networks. <a href="https://app.daily.dev/posts/MtreEtTLc" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>Databricks and Snowflake face market maturation, pivot to AI</strong>: Databricks and Snowflake are confronting growth saturation in their core data platform markets, leading both to aggressively pivot into AI and application development to sustain high valuations, as analyzed by The Great Data Closure. <a href="https://app.daily.dev/feed-by-ids?id=Kv2NZR8R1&amp;id=LnOTpvngc" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Microservices architecture can kill performance</strong>: Microservices introduce significant performance overhead, including 1,000-5,000x latency for network calls compared to in-process functions, leading to higher resource usage, increased costs, and cascading failures, suggesting modular monoliths are often a better choice for most organizations. <a href="https://app.daily.dev/posts/OC7oxM66D" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>ClickHouse acquires Langfuse, launches native Postgres service</strong>: ClickHouse acquired Langfuse, an open-source LLM observability platform, to enhance its AI application monitoring capabilities and launched a native Postgres service in partnership with Ubicloud for unified transactional and analytical workloads. <a href="https://app.daily.dev/feed-by-ids?id=BqVJ6XPkS&amp;id=7DDF3YQda&amp;id=kU2RvmlCw" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>GitHub Copilot introduces agentic memory in public preview</strong>: GitHub Copilot now features agentic memory in public preview for paid plans, allowing it to learn and retain repository-specific insights to improve code assistance, with memories expiring after 28 days and configurable by users and organizations. <a href="https://app.daily.dev/posts/nCmOnsQS3" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Cloudflare launches resilience plan, acquires Astro Technology</strong>: Cloudflare introduced ‘Code Orange: Fail Small,’ a resilience plan with phased configuration rollouts and automatic rollbacks to prevent global outages, and acquired Astro Technology Company, an open-source framework for content-centric websites. <a href="https://app.daily.dev/feed-by-ids?id=OdNHxfcwg&amp;id=Q4Q57JyYL&amp;id=yRAOJZmiR" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Go 1.26 RC2 released with security fixes</strong>: Go 1.26 Release Candidate 2 is out with security fixes for <code>archive/zip</code> and <code>Request.ParseForm</code>, with identical updates also released for Go 1.25.6 and Go 1.24.12. <a href="https://app.daily.dev/posts/DFYbtJ1rQ" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>New open-source database tools: Portabase and credcheck</strong>: Portabase, a new self-hosted backup and restore tool for PostgreSQL, MySQL, and MariaDB, now uses Rust agents for improved performance, while credcheck v4.4, a PostgreSQL extension for credential rule enforcement, fixes a backend crash. <a href="https://app.daily.dev/feed-by-ids?id=5HNbhZxB9&amp;id=7mmyYg2vc" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>OpenAI to introduce targeted ads in ChatGPT: for the users, of course</strong>: OpenAI is graciously bringing targeted ads to free and Go ChatGPT users in the U.S., starting January 16, 2026. They assure us these ads, conveniently placed at the bottom of conversations, will be ‘relevant’ and won’t influence responses. And don’t worry, your data won’t be sold—they just need to ‘support continued free access’ and ‘expand affordable AI access globally.’ Because nothing says ‘affordable AI’ like more ads. <a href="https://app.daily.dev/posts/DQh9IAXNl" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Google’s AI search: ‘unavoidable’ errors, but don’t you dare question it</strong>: Google’s AI search is apparently so advanced it can generate false accusations of criminal behavior against Norwegian politicians, including pedophilia. When confronted, Google’s defense is that such errors are ‘unavoidable’ given the scale of the internet. So, while they push AI as revolutionary, they’re quick to shrug off responsibility for reputational damage, effectively telling users to just deal with the ‘unavoidable’ consequences of their groundbreaking tech. <a href="https://app.daily.dev/feed-by-ids?id=507BvKbPR&amp;id=QiN4er8h3" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI-powered coding tools: still not magic, apparently</strong>: A developer used Claude AI to build a Python CLI tool, cutting development time from a day to an hour. Impressive, right? Except Claude couldn’t handle a robust resume function, requiring manual intervention for file caching. So, while AI can speed up routine scripting, it still falls flat on nuanced requirements and error handling. It seems we’re still a long way from AI magically writing perfect code without human oversight. <a href="https://app.daily.dev/feed-by-ids?id=SqdiiasJQ&amp;id=YfWBhNQB7" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
