<h2>Must know</h2>
<ul>
<li><strong>Google releases powerful new Gemini 3.1 Pro model</strong>: Google just dropped Gemini 3.1 Pro in public preview, touting it as their most powerful agentic and coding model yet. It scored 77.1% on the ARC-AGI-2 benchmark, a massive jump from its predecessor, and features a 1 million token context window. This model is available across various platforms, including GitHub Copilot and Google AI Studio, and is priced competitively against rivals like Anthropic’s Opus 4.6, making it a significant contender in the AI model landscape. <a href="https://app.daily.dev/posts/PjE6Xo1XG" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Anthropic’s Claude Code policy shift causes developer uproar</strong>: Anthropic updated its Claude Code documentation, restricting OAuth token usage from Pro/Max accounts in third-party tools, sparking community backlash. While Anthropic claims it’s just a documentation cleanup, actions like sending a cease-and-desist to OpenClaw suggest a shift towards requiring paid API keys for commercial development. This move signals a potential end to affordable flat-rate AI subscriptions for developers building businesses, pushing them towards more expensive API authentication. <a href="https://app.daily.dev/feed-by-ids?id=3lsZ5Vula&amp;id=FQxtQgg9R&amp;id=GYB8zPRsi" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>MySQL governance under Oracle challenged by community</strong>: Over 248 database professionals, including engineers from major tech companies, have signed an open letter urging Oracle to transfer MySQL governance to an independent non-profit foundation. Concerns include declining commit activity, a shrinking contributor pool, lack of AI-era features, and an opaque roadmap. The community points to PostgreSQL’s success with community-led governance as a model, highlighting growing fragmentation and a desire for more open development. <a href="https://app.daily.dev/posts/GY7vcPtOu" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>AI agents reshape software development workflows</strong>: Agentic AI is transforming software development, with tools like Kiro and Claude Code boosting productivity, but also creating new roles like Agent Orchestration Engineer and raising concerns about bias and compliance. <a href="https://app.daily.dev/feed-by-ids?id=037a5YIjx&amp;id=4CjaWv3pP&amp;id=7XxAvY7Df&amp;id=Bl6akHr3q&amp;id=DqRWRtC6t&amp;id=F8eGl7g0G&amp;id=LCg5ueVO7&amp;id=SBASpAo7A&amp;id=UJqBwwcg1&amp;id=Uw1m3wdti" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI memory and agent management solutions emerge</strong>: New solutions like Mengram’s open-source AI memory API and Reload’s Epic platform are tackling AI agent memory and collaboration challenges, offering semantic, episodic, and procedural memory types for individual agents and shared long-term memory for teams. <a href="https://app.daily.dev/feed-by-ids?id=0tYkAozTe&amp;id=MQxvV1b9u" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Wix engineering experiment shows AI productivity gains</strong>: A 5-week experiment at Wix Engineering found that while some AI coding tools like Cursor showed little measurable productivity lift despite user perception, Claude Code delivered significant output gains once developers built trust, leading to a company-wide rollout. <a href="https://app.daily.dev/posts/qQhctEUPB" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI infrastructure and energy demands surge</strong>: Cisco and NVIDIA are partnering on a Secure AI Factory to meet the growing demand for AI-ready infrastructure, while SoftBank plans a massive $33 billion natural gas power plant to fuel AI data centers, highlighting the immense energy needs of the AI boom. <a href="https://app.daily.dev/feed-by-ids?id=IYsqAR4Cy&amp;id=VBQEBjah6" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Redwood materials expands energy storage for AI data centers</strong>: Redwood Materials, backed by Google and Nvidia, is rapidly expanding its energy storage business unit, repurposing EV batteries to meet the surging power demands of AI data centers, which face significant grid connection delays. <a href="https://app.daily.dev/posts/1mD3jO5pQ" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI and robotics industrial adoption accelerates</strong>: Toyota is deploying Agility Robotics’ Digit humanoid robots in its RAV4 plant for auto parts unloading, while NVIDIA’s survey shows 90% of telecom companies report AI increasing revenue and reducing costs, signaling a broader industrial adoption of AI and robotics. <a href="https://app.daily.dev/feed-by-ids?id=WnnenuOfb&amp;id=p1OVwpWSR" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>OpenAI expands strategic partnerships in India</strong>: OpenAI is deepening its presence in India through partnerships with Pine Labs for fintech integration and Reliance for AI-powered search in JioHotstar, aiming to reduce latency and meet local compliance needs in a market with over 100 million weekly ChatGPT users. <a href="https://app.daily.dev/feed-by-ids?id=JlixRkBbI&amp;id=JodTqUZtB" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>AI’s text-trap: The commoditization of SaaS</strong>: Apparently, text-only AI assistants are going to commoditize SaaS products by stripping away all that ‘carefully crafted user experience.’ Because, you know, nobody wants a simple, efficient text interface when they could have a bloated, over-engineered UI. The solution? More UI components and ‘co-creation’ with the AI. Sounds like a great way to make sure we never actually get anything done, just endlessly co-creating with our digital overlords. <a href="https://app.daily.dev/posts/Nz3KEZcnf" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI code quality and production visibility concerns</strong>: Some folks are worried that AI coding tools are making JavaScript shipping too fast, leading to ‘subtle production bugs’ that traditional code review can’t catch. Their brilliant solution? ‘Production visibility’ and, conveniently, a specific tool called TrackJS. Because the problem isn’t that AI might be generating questionable code; it’s that we’re not paying enough to monitor the mess it makes. Classic. <a href="https://app.daily.dev/feed-by-ids?id=2Xamq3Lpe&amp;id=IR0A3b0XK" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI’s impact on work, authenticity, and multilingual safety</strong>: We’re being told AI replaces tasks, not entire human roles, and that customer acceptance of AI tools is growing. Meanwhile, Microsoft’s AI safety team proposes blueprints for verifying online content authenticity, but the chief scientific officer won’t commit to implementing them on Microsoft’s own platforms. And, surprise, AI safety guardrails are failing in non-English languages. So, AI is great for productivity, but don’t trust it, especially if you don’t speak English, and definitely don’t expect the big players to actually follow their own safety advice. Got it. <a href="https://app.daily.dev/feed-by-ids?id=4HNvjLVYH&amp;id=QMBkqUnaK&amp;id=SCrxseqRQ&amp;id=ZyR7Yr1YW" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
