<h2>Must know</h2>
<ul>
<li><strong>AI solves a previously unsolved math problem</strong>: GPT-5.2 Pro, an AI language model, has reportedly solved Erdos problem #281, a mathematical challenge that human mathematicians had not cracked. Terence Tao, a highly respected mathematician, called this “perhaps the most unambiguous instance” of AI solving an open problem, marking a significant milestone for AI in mathematical research. This achievement, reported by multiple sources, highlights AI’s growing capability in complex problem-solving. <a href="https://app.daily.dev/posts/pRkd3CPEF" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Snap Store publisher accounts hijacked for malware distribution</strong>: Scammers are exploiting the Canonical Snap Store by registering expired domains previously used by legitimate publishers, then taking control of their accounts. This allows them to publish malicious updates to trusted snap applications, specifically targeting cryptocurrency wallet users to steal recovery phrases. This escalation, reported by cybersecurity news, undermines trust in Snap publisher reputation and highlights critical security vulnerabilities in the platform’s verification processes. <a href="https://app.daily.dev/posts/8RJLmL5a0" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>China plans massive satellite constellation, dwarfing Starlink</strong>: China has filed requests with the ITU for two satellite constellations, CTC-1 and CTC-2, totaling an astounding 193,428 satellites—four times the size of SpaceX’s Starlink plans. These filings, made by a newly established Chinese entity, raise national security concerns for the US FCC, which has accelerated approvals for domestic satellite networks in response. While China’s current deployment is limited, this move signals a significant escalation in the global space race, as reported by multiple outlets. <a href="https://app.daily.dev/posts/9Dqo8zul1" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>Physical AI and AV industry shifts at CES 2026</strong>: CES 2026 saw Physical AI, merging AI with real-world control, take center stage, with Mobileye acquiring Mentee Robotics for $900 million and Chinese AV companies increasing their presence amidst US automaker absence, as reported by TechCrunch. <a href="https://app.daily.dev/posts/MtQzzuuIN" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>New CLI tool integrates AI and developers with Figma</strong>: A new command-line tool, dannote/figma-use, offers full read/write access to Figma using JSX syntax, enabling AI agents and developers to programmatically create and modify designs, with an experimental ‘render’ command promising up to 100x faster operations. <a href="https://app.daily.dev/posts/DfAoRLrdU" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Iran disrupts Starlink internet with military-grade GPS jamming</strong>: Iran successfully used military-grade GPS jammers to cut Starlink performance by up to 80% during a nationwide internet shutdown, raising concerns about the satellite internet’s reliability in conflict zones and its perceived censorship resistance. <a href="https://app.daily.dev/posts/VVK9EFJ8w" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Open source models: cathedral, megachurch, and bazaar</strong>: Open source projects are categorized into three models: exclusive ‘cathedrals,’ large ‘megachurches’ like the Linux Foundation, and the vast majority of ‘bazaar’ projects, which are often single-maintainer but critical to infrastructure and face significant funding challenges. <a href="https://app.daily.dev/posts/rr8Vo8FaG" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Python extension is VS Code’s most downloaded ever</strong>: The Python extension for VS Code has hit 200 million downloads, making it the most downloaded extension by far, offering comprehensive features for Python development including IntelliSense and debugging. <a href="https://app.daily.dev/posts/Kt6jX78eK" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>FluentAI: low-code platform for LLM workflows</strong>: fluentAI is an open-source, low-code/no-code platform for visually building custom workflows with large language models in the browser, inspired by Node-RED and built with ReactFlow and Supabase. <a href="https://app.daily.dev/posts/T2Zh09ohU" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Curl drops OpenSSL-QUIC backend for HTTP/3</strong>: The curl project is removing support for the OpenSSL-QUIC backend in version 8.19.0 due to its inadequate API, extremely poor performance (3x slower), and excessive memory usage (20x more), consolidating HTTP/3 support to nghttp2+nghttp3 and quiche. <a href="https://app.daily.dev/posts/As2f3e0AW" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>AI coding agents: the joy killer?</strong>: After 20 years of Vim/Neovim, some developers are finding AI coding agents “good enough” to question if learning fundamental tools like awk and sed still makes sense. It seems the efficiency gains are real, but the existential dread of losing the ‘craft’ that brought joy is also very real. So, is AI making us better, or just making us sadder, more efficient button-pushers? <a href="https://app.daily.dev/posts/vFzyyNK3z" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Soft skills are the new hard skills, thanks to AI</strong>: Apparently, with AI coding agents getting so good at spitting out code, the bottleneck has shifted from writing it to, get this, <em>talking</em> about it. Now, engineers need to master communication and specification skills because AI can’t read your mind. So, while AI writes the code, you still have to deal with the humans—what a concept! <a href="https://app.daily.dev/posts/sifgyKGyd" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>GitHub contribution graphs: still meaningless</strong>: In a shocking turn of events, GitHub contribution graphs, LinkedIn profiles, and LeetCode challenges are still terrible indicators of a developer’s actual worth. Who knew that a few green squares don’t capture the nuance of real-world problem-solving or the context of different job markets? It’s almost like trust and understanding individual circumstances are more important than standardized, easily gamed metrics. <a href="https://app.daily.dev/posts/WQGtqJsDu" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
