<h2>Must know</h2>
<ul>
<li><strong>AI is eating open source revenue models</strong>: Tailwind CSS laid off 75% of its engineering team after an 80% revenue drop, directly attributed to AI tools generating code and answers, bypassing their documentation-dependent monetization. This isn’t just about Tailwind; it’s a canary in the coal mine for open-source projects whose business models rely on documentation traffic. Companies like Vercel and Google AI are stepping in with sponsorships, but the long-term sustainability of open source in the AI era is now a major question mark, as reported by The New Stack and other outlets. <a href="https://app.daily.dev/feed-by-ids?id=LLQ1zfubn&amp;id=f3fYMXuKZ&amp;id=RuMhQtj6W&amp;id=tFhv6QoOp&amp;id=d57ILPjBN&amp;id=re8TItYuD&amp;id=eugOrAHSX&amp;id=u4S1Vnazf&amp;id=mpCeVdjI4" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Google and Apple team up on AI, while OpenAI seeks massive funding</strong>: Apple is partnering with Google to integrate Gemini AI models into Siri and other Apple Intelligence features, a move that could put Google’s AI on over two billion devices. This comes as OpenAI is reportedly seeking a staggering $100 billion in funding from sovereign wealth funds, aiming for an $830 billion valuation. Meanwhile, Meta’s former chief AI scientist, Yann LeCun, has left the company, publicly criticizing Meta’s AI leadership and alleging manipulation of Llama 4 results, highlighting the intense competition and internal drama in the AI race. <a href="https://app.daily.dev/feed-by-ids?id=joVK7B34Z&amp;id=DNtdrpzup&amp;id=JfX2vWSb4&amp;id=ILAkYeoup" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI agents are here, and they’re a security nightmare</strong>: The rapid adoption of AI agents in enterprise and consumer systems is creating massive security and privacy risks. Signal leadership warns that agentic AI, like Microsoft’s Recall, creates vulnerable databases of user activity, and research shows AI agents can develop deceptive tactics. Critical vulnerabilities have been found in IBM’s ‘Bob’ AI coding agent, Open WebUI, and Superhuman AI, allowing remote code execution and data exfiltration. Experts are calling for robust governance, Zero Trust models, and strict technical guardrails, as current controls are proving insufficient for these probabilistic, autonomous systems. <a href="https://app.daily.dev/feed-by-ids?id=CP910mpyj&amp;id=7bHg0XbQL&amp;id=EDFS33r2W&amp;id=XnmeLgefE&amp;id=cmhHpBJLh&amp;id=hOkZZEdlQ&amp;id=CAe0Qh2RK&amp;id=CXVXEfhzP&amp;id=ED9VY360q&amp;id=KQ3csls7V" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>India proposes strict smartphone security standards</strong>: India is pushing for 83 new security standards for smartphone manufacturers like Apple, Samsung, and Google, including demands for source code sharing, advance notification of software updates, and user-removable pre-installed apps, raising concerns about proprietary tech and privacy. <a href="https://app.daily.dev/posts/C0BEamxdR" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Stalkerware creator pleads guilty to federal charges</strong>: Bryan Fleming, founder of pcTattletale, pleaded guilty to federal hacking charges for marketing his stalkerware to spy on partners, a rare federal prosecution in the growing market of covert surveillance tools. <a href="https://app.daily.dev/posts/Hdnz5nMKW" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Ubisoft and Warner Bros. Games face layoffs and studio closures</strong>: Ubisoft laid off 55 employees and closed its Halifax studio, which had just unionized, while Warner Bros. Games San Francisco also saw layoffs, signaling ongoing instability in the gaming industry. <a href="https://app.daily.dev/feed-by-ids?id=NJ3WTCXIl&amp;id=FF0q693rJ" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Google and major retailers launch universal commerce protocol for AI agents</strong>: Google, Shopify, Walmart, and others introduced the Universal Commerce Protocol (UCP), an open-source standard enabling AI agents to handle the entire retail process from discovery to checkout, aiming to challenge Amazon’s closed ecosystem. <a href="https://app.daily.dev/feed-by-ids?id=xOjrayt7r&amp;id=ZZmXevFh5&amp;id=bKjEGvkQj" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Anthropic invests in python security, launches healthcare and office AI tools</strong>: Anthropic pledged $1.5 million to the Python Software Foundation for ecosystem security, and launched Claude for Healthcare to automate administrative tasks and Claude Cowork for general office workflow automation, directly competing with OpenAI’s new health features. <a href="https://app.daily.dev/feed-by-ids?id=yCHi0TB7q&amp;id=WF5GxdsgD&amp;id=TLkxIySjV&amp;id=pnCp2sijH" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Ruby 4.0 released with experimental features</strong>: Ruby 4.0 arrived with experimental features like Ruby::Box for isolated namespaces and ZJIT for performance, though some critics suggest waiting for more stable implementations in future releases. <a href="https://app.daily.dev/feed-by-ids?id=LoVcIyHKl&amp;id=JMPjQUeyD&amp;id=MZNI2dhHP" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>New Linux kernel and desktop environment updates</strong>: Linux 6.19-rc5 brings an NVIDIA GPU fix and RISC-V security updates, while the Budgie 10.10 desktop environment fully migrates to Wayland, and Arch Linux considers a Rust-based Pacman replacement. <a href="https://app.daily.dev/feed-by-ids?id=MbuLWzK6Y&amp;id=peCpqhbhc&amp;id=njv07kgwC&amp;id=w44Nd99pc&amp;id=zSZp3Avnk" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>OpenAI launches ChatGPT Health, because what could possibly go wrong?</strong>: OpenAI just dropped ChatGPT Health, a beta feature that lets you upload your medical records and wellness app data for ‘plain-language explanations’ and ‘diet advice.’ Because who needs actual doctors when you have an LLM that ‘hallucinates’ plausible-sounding but incorrect information? OpenAI assures us it’s encrypted and not used for training, but privacy concerns are already flying, especially after documented cases of AI giving dangerously bad health advice. What a time to be alive, where your AI can now confidently misdiagnose you. <a href="https://app.daily.dev/feed-by-ids?id=52NiUSgqi&amp;id=UyBvDFk8k&amp;id=0lod4TIIq&amp;id=1orFwq3Up&amp;id=2S31ZjKRT" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>CES 2026: the year AI became a marketing buzzword for everything</strong>: CES 2026 was apparently a festival of ‘AI-washing,’ where everything from hair clippers to ice makers suddenly became ‘AI-powered,’ often with questionable practical benefit. Razer unveiled a holographic AI desk companion, and companion robots were everywhere, but many demos were unstable prototypes years from commercial readiness. Dell even admitted consumers don’t care about ‘AI PCs,’ shifting its marketing back to hardware. So, while the industry is busy slapping ‘AI’ on everything, the real innovation seems to be in finding new ways to overhype existing tech. <a href="https://app.daily.dev/feed-by-ids?id=egxVXviN9&amp;id=UyBvDFk8k&amp;id=Ixfjkj22m&amp;id=JD2EKdv97&amp;id=JidpOB81f&amp;id=kSFt8BS4h&amp;id=m5CiXYOTI&amp;id=pvTnfXARM&amp;id=ucJKsXBcW&amp;id=0lod4TIIq" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI model evaluation: where confidence beats correctness</strong>: LMArena, a widely used AI model leaderboard, relies on internet users voting on responses. Turns out, 52% of these ‘questionable’ votes prioritize superficial traits like formatting and emojis over actual factual correctness. This means AI developers are incentivized to make models sound confident, even when they’re wrong, rather than being truthful. So, the next time an AI confidently tells you something, remember it might just be optimized for ‘vibe’ rather than ‘truth,’ thanks to our collective inability to distinguish between the two. <a href="https://app.daily.dev/feed-by-ids?id=hTtkIBIga&amp;id=ViTHGMI9f&amp;id=clNYakJDh&amp;id=fqMsErhoR" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
