<h2>Must know</h2>
<ul>
<li><strong>AI is coming for your job, but not how you think</strong>: New studies suggest AI is already impacting employment, particularly for young workers in AI-exposed roles like software development and customer service. A Stanford-led study, cited by The Register, found a 13% employment decline for 22-25 year olds in these areas since ChatGPT’s launch. This isn’t just about automation; AI hiring systems show a significant self-preference bias, favoring AI-generated resumes over human ones, even with equal qualifications. So, while AI might not replace you directly, it might just prefer someone who used its cousin to write their resume.</li>
<li><strong>OpenAI buys its way into faster product launches</strong>: OpenAI just dropped $1.1 billion on Statsig, a company specializing in A/B testing and product analytics. According to InfoWorld, this move is all about accelerating generative AI product launches and improving iteration cycles for things like ChatGPT. They also snagged the Alex team, known for an AI coding assistant for Xcode, to bolster their Codex team. It seems OpenAI is serious about getting its AI out the door faster, even if it means buying the tools to do it.</li>
<li><strong>Cyber threats are getting spicy with AI</strong>: The cyber landscape is heating up, with AI-fueled threats becoming more sophisticated. Anthropic’s August 2025 report warns of ‘vibe hacking,’ where AI agents autonomously conduct cyberattacks, making advanced operations accessible to even low-skilled criminals. The Register also reports that cybercriminals are already leveraging open-source AI tools like HexStrike AI to weaponize exploits rapidly. Meanwhile, over 1,100 Ollama AI servers were found exposed online due to weak security, proving that accessibility often trumps security in AI deployment.</li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>Android’s big September update for developers</strong>: Google rolled out new Android development tools, including adaptive layouts, AI integration, and Wear OS enhancements, with Android Studio Narwhal featuring Gemini AI and XR development support, and Kotlin 2.2.0 improving multiplatform capabilities.</li>
<li><strong>Go gets a security patch</strong>: Microsoft released security patches for Go 1.25.1 and 1.24.7, addressing a vulnerability in the net/http package that could allow CrossOriginProtection bypass via overly broad URL patterns.</li>
<li><strong>New AI tools for your code editor</strong>: Warp launched ‘Warp Code,’ an agentic development environment for prompt-to-production workflows, while Zed editor integrated Claude Code as a public beta, both aiming to streamline AI-generated code with features like agent steering and real-time edits.</li>
<li><strong>Cloud data platforms get new features</strong>: Google Cloud introduced soft failover for BigQuery Managed Disaster Recovery, allowing safer testing of disaster scenarios, while Databricks made Materialized Views and Streaming Tables generally available via Delta Sharing for real-time data collaboration.</li>
<li><strong><a href="http://Matrix.org" target="_blank" rel="noopener nofollow">Matrix.org</a> homeserver hit by RAID failure</strong>: <a href="http://Matrix.org" target="_blank" rel="noopener nofollow">Matrix.org</a>’s main homeserver experienced an outage due to a RAID failure affecting both primary and secondary databases, leading to a lengthy recovery process for its 55 TB PostgreSQL database and millions of queued messages.</li>
<li><strong>React ecosystem keeps evolving</strong>: The React Compiler and React Native 1.0 received updates, Redux Toolkit 2.9 improved performance, and new tools like BaseUI beta.3 and Reanimated 4.1 were released, while debates continue on React Transitions and Server Components.</li>
<li><strong>African ed-tech platforms are expanding</strong>: African ed-tech platforms like CambriLearn are building localized solutions for intermittent internet, while The Invigilator raised $11 million for its AI-driven exam monitoring solution, both aiming for international expansion.</li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>AI’s cognitive offloading: the new brain drain?</strong>: A study from MIT, reported by the Public Health Policy Journal, claims students using ChatGPT for essay writing showed impaired neural connectivity and memory recall. Apparently, our brains are ‘cognitively offloading’ when AI does the heavy lifting. So, if you feel dumber after using ChatGPT, don’t worry, it’s just science, not your brain turning into mush from too much TikTok.</li>
<li><strong>AI business value: the emperor’s new clothes?</strong>: TowardsDataScience reports that only 5% of AI business initiatives actually deliver meaningful value, mostly due to poor planning and inflated expectations. Meanwhile, Anthropic just secured another $13 billion in funding, pushing its valuation to $183 billion. Investors are still throwing money at AI firms despite reports of 95% of deployments yielding no measurable return. Sounds like the dot-com bubble all over again, but this time with more GPUs and less actual profit.</li>
<li><strong>The death of phones: because we all need more screens on our faces</strong>: UX Planet is predicting the demise of smartphones, replaced by smart glasses, smartwatches, and ambient camera networks that will overlay digital info directly onto reality. And get this: AI agents will handle your social media, purchasing, and even negotiate with brands for you. Because who needs authentic human interaction or the ability to make their own choices when an AI can do it all for you? What could possibly go wrong?</li>
</ul>
