<h2>Must know</h2>
<ul>
<li><strong>NVIDIA and Lilly team up for AI drug discovery</strong>: NVIDIA and pharmaceutical giant Lilly are investing $1 billion in a co-innovation lab to revolutionize drug discovery. This initiative, announced at the J.P. Morgan Healthcare Conference, aims to integrate computational modeling with experimental wet labs, leveraging NVIDIA’s AI infrastructure, including the DGX SuperPOD with B300 systems. The goal is to model biological complexities more accurately and accelerate the discovery of new molecules, as reported by various sources covering the conference. <a href="https://app.daily.dev/posts/TcHQkCPbq" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Microsoft pushes unified AI governance and marketplace expansion</strong>: Microsoft has been recognized as a leader in Unified AI Governance Platforms by IDC MarketScape for 2025-2026, integrating responsible AI standards across its ecosystem. The platform offers centralized oversight for traditional ML, generative AI, and agentic AI, with features like real-time threat detection and automated compliance. Additionally, Microsoft’s Marketplace now boasts over 11,000 AI models and 4,000 AI applications, including pre-vetted solutions from partners like Anthropic, OpenAI, Meta, and NVIDIA, aiming to accelerate enterprise AI adoption with a focus on security and compliance. <a href="https://app.daily.dev/feed-by-ids?id=VuTiPUrsF&amp;id=XdBf9IqBR" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI solves a long-standing Erdos problem</strong>: GPT 5.2 Pro, an AI model, has reportedly solved Erdos problem #281, a previously unsolved mathematical challenge. Mathematician Terence Tao, a highly credible source, stated this is “perhaps the most unambiguous instance” of AI solving an open mathematical problem. This marks a significant milestone for AI in mathematics, particularly in AI-assisted research, signaling a rare clarity in attributing a new mathematical discovery primarily to artificial intelligence. <a href="https://app.daily.dev/posts/pRkd3CPEF" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>NVIDIA expands GeForce NOW to Linux and Fire TV</strong>: NVIDIA announced native GeForce NOW apps for Linux PCs (Ubuntu 24.04+) and Amazon Fire TV devices, enabling high-resolution, high-framerate cloud gaming powered by RTX 5080 GPUs. <a href="https://app.daily.dev/posts/5ceYZHV0v" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>NVIDIA launches DLSS 4.5 and other AI enhancements</strong>: NVIDIA released DLSS 4.5 with Dynamic Multi Frame Generation for RTX 50 Series GPUs, a second-gen transformer model for Super Resolution, and new G-SYNC Pulsar monitors with claimed 1,000Hz+ motion clarity. <a href="https://app.daily.dev/posts/jXjzYY9vI" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>NVIDIA DRIVE powers Mercedes-Benz CLA and expands ecosystem</strong>: NVIDIA DRIVE AV software will be in the new Mercedes-Benz CLA by late 2026, focusing on level 2 driver assistance, while the DRIVE Hyperion ecosystem expands with partners like Bosch and Magna for level 4 autonomous driving. <a href="https://app.daily.dev/feed-by-ids?id=kCsBTg7xZ&amp;id=pTE08FtLn" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI agents in gaming environments show promise and limitations</strong>: Experiments with AI agents in Pokémon Showdown and RollerCoaster Tycoon 2 demonstrate AI’s ability for rule-based reasoning but highlight weaknesses in domain-specific and spatial understanding. <a href="https://app.daily.dev/feed-by-ids?id=6IAZSWIz1&amp;id=NVzHSqkMx" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>KDE Plasma 6.6 nearing release, 6.7 development begins</strong>: KDE developers are finalizing Plasma 6.6 for a February 2026 release while simultaneously starting feature implementation for Plasma 6.7, indicating a productive development cycle. <a href="https://app.daily.dev/posts/Htkecunm3" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>New programming language Kip uses Turkish grammar</strong>: Kip, an experimental programming language, uses Turkish grammatical cases in its type system to define argument relationships in functions, allowing for flexible argument ordering and exploring linguistic morphology in type theory. <a href="https://app.daily.dev/feed-by-ids?id=f4hbmP6tB&amp;id=kTJVodinv" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Porffor JavaScript compiler targets embedded systems</strong>: Porffor is a new ahead-of-time JavaScript compiler aiming for WebAssembly and native code, promising up to 50x smaller binaries and significant speed improvements for serverless, game engines, and embedded systems. <a href="https://app.daily.dev/feed-by-ids?id=xTAvQVuAU&amp;id=xU3ZYAZlY" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>AI agents are just old APIs with new lipstick</strong>: Apparently, the tech world has discovered ‘AI agents,’ which, according to some, are just a fancy new name for the same old API and web automation we’ve had for ages. The hype suggests this is all investor-driven rebranding for financial gain, not some groundbreaking technological leap. So, next time someone talks about ‘agentic automation,’ remember it’s probably just a glorified script with a marketing budget. <a href="https://app.daily.dev/feed-by-ids?id=4rJ96gwiU&amp;id=J1BKrq2u6&amp;id=JE89HgTw1&amp;id=JItr1KihQ&amp;id=bXoqeH8eP" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Claude Co-work: A taste of AGI, if AGI had terrible UX</strong>: Anthropic’s new Claude Co-work desktop app promises to bring Claude Code’s power to the masses via a sandboxed Ubuntu VM. It can organize files and execute terminal commands, which sounds like AGI for your desktop. However, early reports suggest it’s plagued with broken authentication, UI glitches, and confusing permissions. So, while the underlying tech might be powerful, the user experience is apparently a masterclass in how to make advanced AI feel like a beta product from 2005. <a href="https://app.daily.dev/posts/OemgIV2rY" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI-assisted coding: The fast track to technical debt</strong>: Developers are sharing stories of AI-generated code that compiles but lacks product context, creates scattered state management, and introduces subtle bugs only caught during integration. One Google engineer’s claim of Claude Code doing a year’s work in an hour turned out to be a ‘simplified toy version.’ It seems AI is great for boilerplate, but for anything complex, it’s more likely to generate architectural nightmares than production-ready code. <a href="https://app.daily.dev/feed-by-ids?id=6XeVWOBXO&amp;id=STWE5tRXe&amp;id=UAphs9F5B&amp;id=euzGVQGzQ&amp;id=iukrPGTg6&amp;id=m6P4U0J0b&amp;id=wCQfiCXDa" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
