<h2>Must know</h2>
<ul>
<li><strong>AI data centers are sucking up all the power and water</strong>: Hyperscale AI data centers, housing hundreds of thousands of GPUs, are consuming over a gigawatt of electricity each, equivalent to a small city. Over half of this power comes from fossil fuels, leading to significant emissions. These facilities also require massive amounts of water for liquid cooling, causing local water shortages, noise pollution, and rising energy costs for surrounding communities. Meanwhile, next-gen nuclear reactors are being touted as the solution, promising safer, faster deployment to meet this insatiable AI demand, as reported by MIT Technology Review. <a href="https://app.daily.dev/feed-by-ids?id=bf0xB32y7&amp;id=qk040xtMi&amp;id=dJDy4tUxK&amp;id=oJJhQtHHE" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Anthropic and OpenAI are fighting over your coding tools</strong>: Anthropic launched Cowork, an agentic AI tool for non-coders, and simultaneously restricted third-party CLI tools like OpenCode from accessing Claude Pro/Max subscriptions via OAuth. This forces developers to use Anthropic’s more expensive pay-per-use API. OpenAI, seizing the opportunity, immediately announced support for ChatGPT subscriptions within OpenCode and other open-source tools, aiming to gain a competitive edge in the developer tools market, according to ‘Anthropic made a big mistake’. <a href="https://app.daily.dev/feed-by-ids?id=TLkxIySjV&amp;id=8dtXe7P09&amp;id=eRi1VQ5SP" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>BreachForums data leak exposes 324,000 cybercriminals</strong>: BreachForums, a notorious cybercrime marketplace, suffered a data breach exposing 323,986 user records, including usernames, emails, IP addresses, and hashed passwords. The leak, publicly disclosed in January 2026, was allegedly carried out by a hacker disillusioned with the forum’s members. Cybersecurity firm Resecurity confirmed the data’s authenticity, noting it contains enough information for law enforcement to identify and prosecute users who thought they were anonymous. This incident is expected to drive cybercriminals to smaller, invite-only platforms. <a href="https://app.daily.dev/posts/iWBcNtvCe" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>Linux desktop ecosystem is diverging</strong>: The Linux GUI landscape is fragmenting, with IceWM 4.0 continuing X11 support, while Budgie 10.10 fully transitions to Wayland and Budgie 11 plans a rewrite using Qt6 and KDE Frameworks. <a href="https://app.daily.dev/posts/peCpqhbhc" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Laravel gets agentic AI capabilities with Neuron AI SDK</strong>: The Neuron AI Laravel SDK, released January 12, 2026, brings agentic AI to PHP developers, allowing them to create autonomous AI agents using Laravel’s tools and supporting multiple AI providers like Anthropic, OpenAI, and Gemini. <a href="https://app.daily.dev/feed-by-ids?id=4bt4Q0nem&amp;id=QVuznBqAB&amp;id=pfLq9n48m" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Bun v1.3.6 boosts performance and adds new APIs</strong>: Bun v1.3.6 introduces new APIs for tarball creation/extraction and JSONC parsing, adds metafile support to Bun.build, and significantly improves performance for Response.json(), async/await, and Promise.race. <a href="https://app.daily.dev/posts/Wx4br4Zyw" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Oracle’s 2026 Java roadmap focuses on performance and concurrency</strong>: Oracle’s Java team announced 2026 priorities for OpenJDK, including Project Valhalla for memory efficiency, Project Leyden for AOT compilation, and Project Loom for structured concurrency, aiming to boost Java’s speed and scalability. <a href="https://app.daily.dev/posts/gmrJODR16" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>GoBruteforcer botnet targets crypto and Linux servers</strong>: The GoBruteforcer malware is exploiting weak credentials in FTP, MySQL, PostgreSQL, and phpMyAdmin services, specifically targeting those commonly found in LLM-generated tutorials, to attack cryptocurrency projects and Linux servers. <a href="https://app.daily.dev/posts/nO4oxH8nA" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Postman acquires Fern for better API documentation</strong>: Postman acquired Fern, a startup specializing in automated API documentation and SDK generation in nine languages, aiming to reduce developer friction and expand Postman’s full API lifecycle offerings. <a href="https://app.daily.dev/posts/9UqlZV9z7" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Auto-CPUFreq 3.0 enhances Linux battery and CPU control</strong>: Auto-CPUFreq 3.0 for Linux, released January 12, 2026, adds CPU turbo mode override via CLI and GUI, improved battery device configuration, and specific support for ASUS laptop charge thresholds. <a href="https://app.daily.dev/posts/umWMqgJxW" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>AI is making code cheap, but software is still expensive. Shocking!</strong>: Apparently, LLM tools like Claude Code have made generating code ‘trivial,’ but building ‘robust, maintainable software’ remains ‘expensive and complex.’ Who would’ve thought that understanding systems, architecture, and edge cases is still more valuable than just spitting out syntax? It’s almost as if AI can’t replace actual engineering judgment, as this article from an unknown source so profoundly points out. <a href="https://app.daily.dev/posts/BCKlippws" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI-generated code is creating a ‘verification bottleneck.’ What a surprise.</strong>: Developers are now spending more time reviewing unreliable AI-generated code than they save writing it, leading to a ‘verification bottleneck.’ The New Stack reports that 96% of developers don’t trust AI-generated code, yet 72% use it daily. So, AI isn’t shrinking developer toil; it’s just shifting it from writing bad code to fixing AI’s bad code. Progress! <a href="https://app.daily.dev/posts/SVMT3qR0j" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Anthropic launches ‘Claude for Healthcare’ to automate admin tasks. What could go wrong?</strong>: Anthropic just launched ‘Claude for Healthcare,’ promising to automate administrative tasks like prior authorizations and claims appeals. They’re very careful to say it’s ‘not for direct clinical decision-making,’ which is reassuring, I guess. It’s HIPAA-compliant and won’t use your data to train models, they claim. Because nothing says ‘trustworthy’ like an AI company promising not to mess with your health data, right? <a href="https://app.daily.dev/posts/WF5GxdsgD" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
