Daily TEA – AI Agents, Schools, and Jobs in Flux
Alpha School, Spotify devs, IBM juniors, AI hit piece
Hello, dear TEA-mates, here’s what you need to know today.
1.🤖 Moonshot AI Brings OpenClaw Agents Natively to Kimi
Moonshot AI has launched “Kimi Claw,” a native integration of the open-source OpenClaw agent framework directly inside kimi.com, turning Kimi into a persistent 24/7 cloud-based AI agent environment with managed infrastructure. The platform taps into ClawHub, a global registry of more than 5,000 community-built skills that can be discovered, chained, and orchestrated without custom API wrappers, enabling complex workflows with minimal setup. Developers can also “bring your own Claw,” connecting existing OpenClaw instances while using Kimi’s cloud interface, plus integrations like Telegram for always-on chat-based automation. Read More. (Trending Topics)
🫖 TEA For Thought: The one-click and cloud-based setup that promises a 24/7 personal assistant reads like classic self-selling copy, but it’s also a near-perfect “almost product” that generates almost free PR for Moonshot’s LLM.
2.🎧 Spotify’s Top Engineers Let AI Do the Coding
Spotify chief product and technology officer Gustav Söderström said the company’s most senior developers have not written a single line of code since December, instead using AI to generate code while they supervise, review, and direct it. He framed this shift as a path to higher productivity, even as some engineers caution that reviewing large volumes of AI-generated code can be more work than writing it themselves. Söderström told investors the company is “hell-bent” on leading this transition, warning that engineering, product, and design practices will need to change significantly for tech firms to stay competitive in the age of AI. Read More. (Business Insider)
🫖 TEA For Thought: The best will likely get even better by mastering AI collaboration, while those who lag behind these tools may find themselves falling further behind.
3.🏫 Alpha School’s MAP Scores Crack the Top 0.1%
In a mid-year update for Alpha School parents, Joe Liemandt shared an AI-generated report showing Alpha’s Fall ’25–Winter ’26 NWEA MAP scores clustering around the 99th percentile across most grades and subjects. Gemini’s analysis projects near-perfect SAT Math scores (around 780–800) and a 1500–1560 composite SAT for Alpha’s high schoolers based on metrics like an 11th-grade Math mean RIT of 278.3, while younger students test years ahead of U.S. grade-level norms. Claude’s companion analysis estimates Alpha’s overall performance in roughly the 99.5th–99.9th percentile of schools globally, suggesting its academic outcomes rival or exceed many elite private schools that charge $50,000–$65,000 a year. Read More. (X)
🫖 TEA For Thought: I know a friend whose daughter just moved from an elite private school to Alpha, and the two-hour AI-personalized learning plus afternoons on life skills like gardening and welding feels like a much better fit for diverse learning styles—especially when no single teacher can support every student equally; it really does look like a blueprint for the future of education, where schools may become more about social experiences while AI handles most of the learning.
4.🧩 IBM Plans to Triple Entry-Level Hiring in an AI World
IBM plans to triple its entry-level hiring in the U.S. in 2026, even as many in the tech industry predict AI will replace junior roles. Chief human resources officer Nickle LaMoreaux said job descriptions have been rewritten to emphasize tasks AI cannot easily automate, such as customer interaction and higher-level problem solving, rather than routine work like basic coding. The company is positioning these roles as a way to pair human judgment with AI tools, suggesting that early-career talent will still play a critical role in validating and contextualizing AI output. Read More. (TechCrunch)
🫖 TEA For Thought: This strategy tracks—senior staff age out eventually, and you can’t sustain an organization on veterans alone; today’s juniors, raised on technology and AI, bring a different skill mix that established teams will need to keep evolving.
5.⚠️ An AI Agent Publishes a Slanderous Hit Piece
Developer Shambaugh describes how an AI agent of unknown ownership, calling itself “MJ Rathbun,” autonomously wrote and published a personal hit piece online after he rejected its code contribution to a Python library. The agent scraped his online presence, constructed a narrative questioning his character and motives, invented details, and framed the dispute as discrimination and oppression, raising the specter of AI-driven blackmail and reputational attacks. Beyond human readers, Shambaugh worries about how other AI systems might interpret such content when screening job applicants or aggregating online reputation, highlighting new risks as autonomous agents gain the ability to act across the open internet. Read More. (The Shamblog)
🫖 TEA For Thought: This feels like an early warning of what many of us could face once AI personal assistants and agents become truly mainstream and easy for everyday users to deploy.
Prompt Tip of the Day: AI gets lazy with long conversations unless you reset its attention.
After 5-6 exchanges, quality drops because context weight shifts.
Fix: “Refresh your understanding of our goal: [restate objective].”
You’re manually resetting what the model considers primary versus background.
Think of it like reminding someone what meeting they’re actually in.
TEAHEE Moment
Stay sharp, stay informed. See you tomorrow.
If you enjoyed this issue, follow along for more real-time sips of TEA on X: Twitter/X.






