Daily TEA – Health, Humanity, Bitcoin, AI Playbooks, and China’s Chip Gambit

Health, Copilot, Reddit badges, Bitcoin reserves, AI adoption, Nvidia, China

Dec 11, 2025

Hello, dear TEA-mates—here’s what you need to know today.

1. 📱 Copilot Data Shows Health Tops Mobile AI Use

Microsoft’s Copilot Usage Report 2025 analyzed 37.5 million de-identified conversations and found that health-related topics are the dominant use case on mobile, consistently ranking as the most common theme across days, months, and times of day. The report highlights patterns like users turning to Copilot for wellness tracking, health tips, and daily routines, as well as spikes in relationship and personal growth queries around Valentine’s Day, late-night questions about religion and philosophy, and a growing shift from pure information lookup to advice-seeking. These trends underscore Copilot’s role as a “trusted companion” embedded in everyday life, driving Microsoft to emphasize privacy-preserving analytics and higher quality standards for responses. (Read More)

🫖 TEA For Thought: Health really is everyone’s top priority—without it, no amount of zeros in your bank account matters, and this report shows AI is now woven into real human life, not just work or study.

2. 🧾 Reddit Pilots Human-Style Verification Badges

Reddit has begun a limited test of verified profiles that place a grey checkmark next to notable people and businesses, aiming to confirm identity rather than sell status. The opt-in badges are meant to help users and moderators know when they’re engaging with experts, journalists, brands, or public figures in moments when verification matters, without changing Reddit’s pseudonymous culture or granting extra privileges. For now, Reddit is manually approving eligible accounts in good standing and excluding NSFW-focused profiles, with plans to move to a third-party verification process as the program scales. (Read More)

🫖 TEA For Thought: In the AI era, proof of humanhood is everything—zero-knowledge proof systems on blockchains may end up being the backbone of telling real people apart from bots, and this shift is pushing the internet toward that zk-proof future faster than ever.

3. ₿ Bitcoin Leaves Exchanges as Reserves Drop 400K BTC

On-chain analytics firm Santiment reports that bitcoin balances on centralized exchanges have fallen by more than 403,000 BTC over the past year, roughly 2% of total supply, as tracked on its Sanbase dashboard. CoinGlass data shows exchanges now hold about 2.11 million BTC, with much of the outflow moving into cold storage and regulated products like ETFs, a pattern Santiment and industry analysts describe as supportive of long-term price stability and a sign of increasingly institutional ownership. Bitcoin has held near the $90,000 level during this period, with reduced exchange inventory historically linked to fewer large, abrupt sell-offs. (Read More)

🫖 TEA For Thought: This trend reinforces how illiquid Bitcoin can be, especially when high fees and slow transaction finality make it harder to move quickly in and out of positions.

4. 🧠 Will Larson Outlines Practical Playbook for Company AI Adoption

Will Larson’s “Facilitating AI adoption at Imprint” lays out a pragmatic strategy for rolling out LLM tools and agents across a company, focusing on iterative experimentation, leadership engagement, and removing friction to everyday usage. The playbook emphasizes auto-provisioned access to a standard AI platform (like ChatGPT or Claude), documenting internal tips and prompts, and partnering closely with teams to embed AI into workflows such as customer support, ticket triage, and compliance Q&A rather than waiting for a “platform” to magically drive adoption. Larson argues that meaningful AI adoption is less about flashy vision decks and more about unglamorous details like account provisioning, access controls, and tailoring approaches for each function so AI becomes a default tool across the organization. (Read More)

🫖 TEA For Thought: This is a very practical guide if you’re building with a team and need a concrete blueprint for rolling out AI beyond slideware.

5. 💾 China Pushes Domestic AI Chips as Trump Opens Door to Nvidia H200 Exports

China has, for the first time, added domestically produced AI chips from groups including Huawei and Cambricon to an official government procurement list, signaling a major effort to boost local semiconductor suppliers in the public sector and reduce reliance on US technology. The Ministry of Industry and Information Technology’s guidance, quietly circulated to government agencies and state-owned enterprises, could translate into billions of dollars of demand for Chinese chipmakers as part of a broader “Xinchuang” strategy that has already begun phasing out foreign processors and operating systems in offices, schools, hospitals, and SOEs. The move came just before President Donald Trump said he would lift US export controls and allow Nvidia to sell its H200 AI chips to “approved customers in China” in exchange for a 25% fee, though those sales still face resistance from US lawmakers and Beijing’s own emerging plans to limit access to imported GPUs and push firms toward domestic alternatives. (Read More)

🫖 TEA For Thought: Chips are now core AI data infrastructure, and Beijing’s latest policy shows both a real push for tech independence and the painful transition costs—companies sitting on idle local GPUs while their Nvidia-based stacks still run the real work.