80 LEVEL

Appcoda

CSS Weekly

Electron

Hacker News

iOS Dev Weekly

<h2>Must know</h2>
<ul>
<li><strong>TSMC sees endless AI chip demand</strong>: TSMC reported record Q4 earnings for 2025, with revenue up 25.5% year-over-year, driven by what CEO C.C. Wei calls an “endless” demand for AI chips. The company forecasts 30% revenue growth for 2026 and plans $52-56 billion in capital expenditures, primarily for advanced 7nm and smaller process nodes. TSMC expects its 2nm manufacturing process to ramp up faster than previous nodes, with about 30% of that production slated for the USA, as reported by TSMC’s official financial statements. <a href="https://app.daily.dev/posts/aTc9ls9yL" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>OpenAI to test targeted ads in ChatGPT</strong>: OpenAI is rolling out targeted ads for free and Go tier ChatGPT users in the U.S. starting January 16, 2026. These ads will appear at the bottom of conversations, tailored to the topic discussed, but OpenAI assures they won’t influence chatbot responses or sell user data. This move, according to OpenAI, aims to support free access to ChatGPT and generate revenue, while Pro, Business, Enterprise, and Plus tiers remain ad-free. <a href="https://app.daily.dev/posts/DQh9IAXNl" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Elon Musk’s lawsuit against OpenAI and Microsoft heads to jury trial</strong>: A federal judge has rejected motions to dismiss Elon Musk’s lawsuit against OpenAI and Microsoft, clearing the way for a jury trial in late April. Musk alleges OpenAI violated its founding nonprofit mission by prioritizing profits after receiving billions from Microsoft, claiming Microsoft knowingly aided in this breach. The jury will now decide if OpenAI broke its promises and if Microsoft assisted in those alleged breaches. <a href="https://app.daily.dev/posts/pwpD2QRjm" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>GitHub Copilot gets agentic memory</strong>: GitHub Copilot now features agentic memory in public preview for all paid plans, allowing it to learn and retain repository-specific insights to improve code assistance, with memories expiring after 28 days and managed by users and admins. <a href="https://app.daily.dev/posts/nCmOnsQS3" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>ClickHouse acquires Langfuse and launches Postgres service</strong>: ClickHouse acquired Langfuse, an open-source LLM observability platform, to enhance its AI application monitoring, and launched a native Postgres service with Ubicloud for unified transactional and analytical workloads. <a href="https://app.daily.dev/feed-by-ids?id=7DDF3YQda&amp;id=kU2RvmlCw&amp;id=BqVJ6XPkS&amp;id=Ycyvr9ct2" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Google previews BigQuery AI for natural language to SQL</strong>: Google is testing a new BigQuery feature called Comments to SQL, powered by Gemini, which translates natural-language instructions within SQL comment blocks into executable SQL queries to accelerate data analysis. <a href="https://app.daily.dev/posts/5AcBUpc2i" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Modular DS WordPress plugin critical vulnerability</strong>: A critical privilege escalation vulnerability (CVE-2026-23550) in Modular DS WordPress plugin versions 2.5.1 and below allows unauthenticated attackers to gain instant admin access, with a patch (v2.5.2) now available. <a href="https://app.daily.dev/posts/D6wKV3WK1" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Palo Alto Networks patches denial-of-service flaw</strong>: Palo Alto Networks released patches for CVE-2026-0227, a high-severity denial-of-service vulnerability in PAN-OS firewalls with GlobalProtect enabled, which could allow unauthenticated attackers to force devices into maintenance mode. <a href="https://app.daily.dev/posts/MtreEtTLc" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Flutter refactors for modularity</strong>: Flutter is refactoring its core SDK to decouple Material and Cupertino design systems, aiming for independent versioning, leaner applications, and a more universal rendering layer. <a href="https://app.daily.dev/posts/T07WzSMOC" 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>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>AI-powered coding tools: the new productivity paradox, again</strong>: Apparently, AI coding tools are so amazing they can cut development time from a day to an hour, as one developer gushes about building a Python CLI tool with Claude AI. Yet, the same developer had to manually implement a ‘robust resume function’ because the AI couldn’t handle it. So, it’s a miracle worker, but only for the easy stuff, leaving us mere mortals to clean up the ‘nuanced software requirements.’ Progress! <a href="https://app.daily.dev/feed-by-ids?id=SqdiiasJQ&amp;id=YfWBhNQB7&amp;id=H6Mp947gD&amp;id=h6ZJabZ0D&amp;id=laMxMdAYV" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>OpenAI’s brain-computer interface: because losing $74 billion by 2028 isn’t enough</strong>: Sam Altman’s Merge Labs just raised $252 million for non-invasive brain-computer interfaces, with OpenAI chipping in despite projections of losing $74 billion by 2028. OpenAI, with its $1.4 trillion infrastructure commitments, is clearly thinking long-term, like ‘decades’ for product commercialization. Because when you’re bleeding money, the obvious solution is to invest in sci-fi tech that might pay off when your grandkids are old. <a href="https://app.daily.dev/posts/rYkFwNLz3" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Cursor AI’s ‘success’ without evidence: the emperor’s new browser</strong>: Cursor published a blog post claiming their autonomous AI agents built a web browser with over a million lines of code in a week, touting ‘meaningful progress.’ However, the code doesn’t compile, has 34 errors and 94 warnings, and there’s no working demo or build instructions. It’s almost as if ‘AI-generated’ now means ‘looks good in screenshots, but don’t actually try to run it.’ <a href="https://app.daily.dev/posts/V3KHuMgUU" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
