<h2>Must know</h2>
<ul>
<li><strong>AI is killing open source business models</strong>: AI tools are causing a revenue crisis for open-source projects, as developers increasingly use AI to get answers and generate code, bypassing official documentation and paid resources. Tailwind CSS, for example, laid off 75% of its engineering team after an 80% revenue drop, despite its framework’s growing popularity, according to Socket and Web Dev Cody. This trend highlights a fundamental challenge for open-source sustainability: AI consumes free content without compensating creators, forcing a shift from content-driven monetization to service-driven models like hosted platforms and APIs. <a href="https://app.daily.dev/feed-by-ids?id=LLQ1zfubn&amp;id=f3fYMXuKZ&amp;id=RuMhQtj6W&amp;id=d57ILPjBN&amp;id=re8TItYuD&amp;id=GEG8xCrY5&amp;id=3KrG8AfuZ&amp;id=WagqERKHJ&amp;id=hsznY6oi4&amp;id=eugOrAHSX" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Microsoft and Google partner for AI, Apple uses Gemini for Siri</strong>: Apple is integrating Google’s Gemini AI models into its core AI features, including a revamped Siri, with a rollout expected this year, as reported by TechCrunch. This partnership, potentially worth $1 billion annually, positions Gemini as the foundation for Apple Intelligence across billions of devices. Meanwhile, Microsoft acquired Osmos, an AI data engineering startup, to automate data preparation within its Fabric platform, aiming to reduce engineering bottlenecks. These moves signal a major consolidation in the AI ecosystem, with tech giants leveraging partnerships and acquisitions to dominate the AI infrastructure race. <a href="https://app.daily.dev/feed-by-ids?id=joVK7B34Z&amp;id=dRdzpOrJb&amp;id=dYYHmjuRB&amp;id=rUWLw3FP" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>OpenAI launches ChatGPT Health, raising privacy concerns</strong>: OpenAI launched ChatGPT Health, a new feature allowing users to upload medical records and connect wellness apps for AI-driven health information. While OpenAI claims conversations are encrypted and not used for model training, privacy and accuracy concerns are widely cited, especially regarding potential misinformation and hallucinations. This launch, along with Anthropic’s Claude for Healthcare, marks a competitive push into AI-driven health solutions, but critics question the sufficiency of privacy measures and the risks of users treating AI as a medical authority. <a href="https://app.daily.dev/feed-by-ids?id=52NiUSgqi&amp;id=WF5GxdsgD" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>Laravel 12.45 adds JSON:API compliance and other features</strong>: Laravel 12.45 introduced JsonApiResource for JSON:API-compliant responses, along with new features like AsBinary cast for binary data, fromCloudStorage() helper for email attachments, and extended support for BackedEnums in session and cache keys. <a href="https://app.daily.dev/posts/Se1yYfwcA" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Anthropic invests $1.5 million in Python security</strong>: Anthropic pledged $1.5 million to the Python Software Foundation to enhance the security of the Python ecosystem, focusing on CPython and PyPI with automated malware detection and security tool development. <a href="https://app.daily.dev/posts/yCHi0TB7q" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI coding agents reshape developer workflows</strong>: AI coding agents like Claude Code are transforming software development, enabling non-programmers to create applications and shifting developers’ focus from writing code to reviewing, integrating, and managing AI-generated output. <a href="https://app.daily.dev/feed-by-ids?id=T7lz8DVGQ&amp;id=ggxtFI8fz&amp;id=zZPpIk0MP&amp;id=FeqXX8Qki&amp;id=QkdkxdhUj&amp;id=QqckxdhUj&amp;id=TaETvLfgQ&amp;id=QJSZXW8qJ&amp;id=WtQOA9MEd&amp;id=f8luCQfKk" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Node.js releases address multiple CVEs</strong>: Node.js released versions 20.20.0 LTS and 25.3.0, addressing multiple CVEs, including fixes for futimes, TLSSocket error handling, and symlink API access permissions. <a href="https://app.daily.dev/posts/DwsoMgMKH" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>New Ruby 4.0 features and stability concerns</strong>: Ruby 4.0, released on its 30th anniversary, introduces experimental features like Ruby::Box for isolated namespaces and ZJIT as a new JIT compiler, though some sources raise stability concerns for production use. <a href="https://app.daily.dev/feed-by-ids?id=LoVcIyHKl&amp;id=JMPjQUeyD&amp;id=MZNI2dhHP&amp;id=UHRxELagh" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Databricks launches Instructed Retriever for enterprise AI</strong>: Databricks introduced Instructed Retriever, an AI retrieval architecture that integrates database queries with RAG similarity search, claiming 35-50% higher retrieval recall and improved accuracy for enterprise AI applications. <a href="https://app.daily.dev/posts/YccZnqmwj" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI agents reshape cloud and data workflows</strong>: AI agents are driving a shift from stateless cloud architectures to persistent, stateful environments, with companies like Databricks and Snowflake deploying AI assistants for conversational data access and debugging, while <a href="http://Fly.io" target="_blank" rel="noopener nofollow">Fly.io</a> launched Sprites for persistent VM sandboxes. <a href="https://app.daily.dev/feed-by-ids?id=2XKWuVI4o&amp;id=OcN7tvK3U&amp;id=SXn4AkZZF&amp;id=xM76X8yai&amp;id=xdHmgnzKW&amp;id=HAEIV3Yfw" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>OpenAI’s ChatGPT Health: Your medical data, their encryption keys</strong>: OpenAI, in its infinite wisdom, has launched ChatGPT Health, where you can upload your most sensitive medical data. They promise it’s encrypted and not used for training, but guess who holds the encryption keys? That’s right, them. Because nothing says ‘privacy’ like handing over your health history to a company whose LLMs are known to ‘hallucinate’ and whose terms explicitly state it’s not for medical diagnosis. What could possibly go wrong? <a href="https://app.daily.dev/posts/52NiUSgqi" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>CES 2026: AI-washing everything from hair clippers to ice makers</strong>: CES 2026 was a glorious parade of ‘AI-powered’ gadgets, from noise-reducing ice makers to ultrasonic chef’s knives. Apparently, every appliance now needs AI, often with questionable practical benefit, according to The Verge. It’s less about innovation and more about slapping ‘AI’ on anything to justify a higher price tag. Because who doesn’t need a holographic AI desk companion that uses Grok AI to tell you what you already know? <a href="https://app.daily.dev/feed-by-ids?id=egxVXviN9&amp;id=UyBvDFk8k&amp;id=Ixfjkj22m&amp;id=JD2EKdv97&amp;id=JidpOB81f&amp;id=kSFt8BS4h&amp;id=m5CiXYOTI&amp;id=pvTnfXARM&amp;id=ucJKsXBcW&amp;id=0lod4TIIq" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI coding assistants: The new productivity paradox</strong>: Developers are reporting that AI coding assistants, while seemingly boosting productivity, actually lead to ‘dead zones’ of 5-30 seconds that fragment focus and invite context switching. Studies even suggest developers take longer to finish tasks despite perceiving a productivity gain. So, while AI writes 70% of the code, you’re still stuck babysitting it, reviewing its ‘hallucinations,’ and dealing with 1.7x more issues. Sounds like a real time-saver, doesn’t it? <a href="https://app.daily.dev/feed-by-ids?id=eC9HsHF69&amp;id=infazu71Y&amp;id=nVFL2FYYR&amp;id=uTQFFroPQ&amp;id=02YtNeXqK&amp;id=1olO8eENO&amp;id=fLWmEtnzQ&amp;id=x6ymvhMnL&amp;id=zpPmuG4cG" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
