<h2>Must know</h2>
<ul>
<li><strong>AI models are getting smarter, but also riskier</strong>: Google has integrated Agentic Vision into Gemini 3 Flash, allowing it to interactively analyze images and execute code for tasks like interface monitoring and robotics automation. Meanwhile, Moonshot AI’s Kimi K2.5, an open-source multimodal model, claims to outperform Gemini 3 Pro and GPT 5.2 on coding and video tasks, leveraging an agent swarm paradigm. However, the increasing sophistication of AI agents also brings new threats, as evidenced by a novel phishing attack using DeepSeek and Gemini to generate polymorphic JavaScript malware, bypassing traditional detection by reassembling client-side, as reported by The Guardian. <a href="https://app.daily.dev/feed-by-ids?id=KIgS0LnaZ&amp;id=6Szzft8Kv&amp;id=jfmiKNwaJ" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Quantum computing advances and a major contract for D-Wave</strong>: D-Wave announced significant progress in both annealing and gate-model quantum computing, with a 314% year-over-year growth in Advantage2 system usage. They also secured a $10 million, two-year Quantum Computing as a Service (QCaaS) contract with an unnamed Fortune 100 company, aiming to develop quantum-powered enterprise applications. Separately, D-Wave, Davidson Technologies, and Anduril demonstrated a quantum-classical hybrid system that accelerated missile defense planning tenfold, improving threat mitigation by 9-12% in simulations, according to company statements. <a href="https://app.daily.dev/feed-by-ids?id=B9UXF3b8Z&amp;id=7x58WcvLH" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Microsoft expands quantum development and AI initiatives</strong>: Microsoft updated its Quantum Development Kit (QDK) with GitHub Copilot integration, specialized chemistry toolkits, and support for multiple quantum languages, aiming to broaden access for developers. Concurrently, Microsoft launched the ‘Frontier Transformation’ AI initiative for enterprises, focusing on embedding AI via Copilots and agents, promoting innovation, and ensuring observability with tools like Agent 365. This initiative highlights implementations by major companies like Epic, Levi Strauss, and Mercedes-Benz, showcasing AI’s growing role in diverse sectors. <a href="https://app.daily.dev/feed-by-ids?id=0i847vEah&amp;id=YPkEwGwyV" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>Google rolls out AI Plus plan globally</strong>: Google’s AI Plus plan, priced at $7.99/month in the U.S. and lower in other regions, is now available globally, offering Gemini 3 Pro, Nano Banana Pro, AI filmmaking tools, NotebookLM, and 200GB storage, directly competing with OpenAI’s ChatGPT Go. <a href="https://app.daily.dev/posts/r06f1pqYU" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>WhatsApp enhances security and adopts Rust</strong>: WhatsApp launched ‘Strict Account Settings’ for high-risk users, blocking media from unknown senders and silencing calls, while also replacing 160,000 lines of C++ code with Rust for its media handling stack to improve security and performance. <a href="https://app.daily.dev/feed-by-ids?id=0u5TXEYDl&amp;id=2YIPmtHiF&amp;id=Fw1AO3ne7" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Linux kernel formalizes succession plan</strong>: The Linux kernel project has formalized a succession plan for leadership after Linus Torvalds’ eventual departure, ensuring continuity and addressing concerns about an aging maintainer community. <a href="https://app.daily.dev/posts/vAKZlS9Qz" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Rust Foundation outlines 2025 annual report and strategic plan</strong>: The Rust Foundation released its 2025 annual report and 2026-2028 strategic plan, prioritizing stable infrastructure, maintainer support, responsible adoption growth, and community engagement, having raised $5.1 million in 2025. <a href="https://app.daily.dev/posts/DRpplNBlF" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Pinterest shifts to AI, announces layoffs</strong>: Pinterest is laying off less than 15% of its workforce (around 700 employees) by late September 2026 to reallocate resources towards AI-focused roles and initiatives, emphasizing AI-powered products like Pinterest Assistant. <a href="https://app.daily.dev/posts/yPKU3lKW1" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Keyfactor and IBM Consulting partner on post-quantum cryptography</strong>: Keyfactor and IBM Consulting have partnered to accelerate enterprise adoption of post-quantum cryptography (PQC), integrating Keyfactor’s PKI tools into IBM’s cybersecurity offerings to defend against future quantum attacks. <a href="https://app.daily.dev/posts/mI2Tee5Vn" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>EU consults on open source strategy</strong>: The European Commission is seeking feedback on its ‘Towards European Open Digital Ecosystems’ strategy, aiming to establish a framework and funding to scale EU open-source projects and strengthen technological sovereignty. <a href="https://app.daily.dev/posts/IteNH1GsU" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>ChatGPT caught citing AI-generated encyclopedia</strong>: The Guardian found ChatGPT’s GPT-5.2 model citing Grokipedia, an AI-generated encyclopedia by Elon Musk, raising concerns about recursive AI content amplifying bias and misinformation. OpenAI claims its models filter low-credibility sources, but apparently, Grokipedia slipped through. So, we’re building AI that learns from other AI’s hallucinations. What could possibly go wrong? <a href="https://app.daily.dev/posts/fWRmbZwbT" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI coding agents: fast, but at what cost?</strong>: While Andrej Karpathy claims to have shifted to 80% AI agent-assisted coding, praising their ‘relentless persistence,’ others are finding that AI agents generate code quickly but often disregard internal quality, leading to ‘spaghetti code’ and subtle semantic errors. Apparently, the ‘AI will replace developers’ hype is still just that, hype, as human oversight remains ‘essential and irreplaceable’ for actual production-level code. So much for the AI revolution making us all obsolete. <a href="https://app.daily.dev/feed-by-ids?id=vRs04mdPH&amp;id=uilUGnhp4&amp;id=fPZ532s1X" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Cloudflare’s ‘serverless’ Matrix homeserver gets roasted</strong>: Cloudflare proudly announced a ‘serverless’ Matrix homeserver on its Workers platform, claiming zero operational overhead and post-quantum security. However, actual Matrix developers are calling it ‘fundamentally broken,’ lacking basic security features and interoperability, with code that looks ‘AI-generated.’ It seems Cloudflare’s marketing department is working overtime, while their engineers might be cutting a few too many corners. Maybe they should stick to what they’re good at, like DDoS protection, instead of reinventing the wheel, badly. <a href="https://app.daily.dev/feed-by-ids?id=vURH9h7xw&amp;id=93HDTbjA4" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
