In Today’s Issue:

🚨 Sam Altman halts secondary projects to focus entirely on Google’s Gemini 3 threat

🏛️ Beijing’s next five-year plan targets 70% AI penetration by 2027

📉 A new report reveals that 21% of YouTube uploads are now low-quality AI-generated content

🚀 The new WeDLM-8B launches on Hugging Face

✨ And more AI goodness…

Dear Readers,

China is quietly rewiring the AI race, its 2026 five-year plan aims to turn AI into everyday infrastructure in factories, schools, hospitals, government, and consumer life, while tightening rules for human-like systems. Around that pivot, today’s edition spotlights Tencent’s new open instruction model and why the sub-10B tier suddenly matters, Sam Altman’s “code red” strategy at OpenAI, the rise of AI “slop” flooding platforms and eroding trust, and a genuinely elegant robotics idea: spiral soft grippers that could make physical automation far cheaper and simpler. Dive in, because the real battle now isn’t just for better models, but for smarter deployment, credible governance, and your scarce attention.

All the best,

Tencent has released WeDLM-8B-Instruct, an 8-billion-parameter open instruction-tuned language model designed for strong reasoning, dialogue, and task-following performance. Hosted on Hugging Face, the model targets developers who want a compact yet capable alternative to larger LLMs, with solid multilingual support and practical deployment efficiency. This signals Tencent’s push into the open-model ecosystem, raising competition in the sub-10B parameter tier where cost-performance matters most.

What are we scaling?

China is preparing to make AI boring, in the best possible way. As Beijing readies its next five-year plan, due in March 2026, officials are signaling a shift away from chasing the flashiest frontier breakthroughs and toward rapid, nationwide deployment of existing models. The stated trajectory is staged: broad AI use across six priority domains by 2027 (research, industrial processes, consumer products, healthcare and education, digital government, and technology exports), then AI becoming as ubiquitous as electricity or the internet by 2030, and an “intelligent society” vision by 2035.

This isn’t just rhetoric. China’s “AI+” policy push frames adoption targets in practical terms, think AI agents and “intelligent terminals” penetrating daily workflows at scale, backed by public-sector coordination and industrial policy. At the same time, regulators are tightening oversight of human-like and emotionally interactive AI services, underscoring that diffusion will come with guardrails.

Why it matters: In AI, diffusion can beat invention, fast rollout turns “good enough” models into massive productivity gains. The next phase of the AI race may be decided in factories, schools, and government services, not benchmark charts.

Sources:
🔗https://www.economist.com/the-world-ahead/2025/11/12/china-will-run-a-different-ai-race-in-the-coming-year
🔗https://english.www.gov.cn/policies/latestreleases/202508/27/content_WS68ae7976c6d0868f4e8f51a0.html
🔗 https://www.reuters.com/technology/china-says-it-will-increase-support-ai-science-tech-innovation-2025-03-05/

Soft robots usually face a brutal trade-off: either they’re gentle but weak, or strong but fussy to build and control. A new open-access paper in Device proposes a surprisingly elegant “escape hatch”: design the body like a logarithmic spiral, the same geometry you see echoed in nature’s best grippers, from octopus arms to elephant trunks.

The team’s “SpiRobs” are 3D-printed from TPU (a flexible plastic) and moved with just two or three cables, like tendons. The result is a soft manipulator that can curl, wrap, and lock onto objects with minimal hardware. Then scale that same idea from centimeters to meters, including multi-robot arrays that can “tangle up” irregular items. In tests, the concept is shown grasping objects that vary by more than two orders of magnitude in size and supporting loads far above its own weight.

If we can standardize a geometry that works across scales, soft robotics may finally get its “LEGO brick.” What would you build if your gripper design scaled as easily as your code?

A universal, cheap-to-fabricate soft gripper can unlock real-world automation in messy environments, warehouses, agriculture, field robotics, where rigid arms still struggle. The big win is deployment speed: fewer custom parts, faster iteration, lower costs.

Sure, if you held your stocks since the dotcom bubble, you would’ve been up—eventually. But three years after the dot-com bust the S&P 500 was still far down from its peak. So, how else can you invest when almost every market is tied to stocks?

Lo and behold, billionaires have an alternative way to diversify: allocate to a physical asset class that outpaced the S&P by 15% from 1995 to 2025, with almost no correlation to equities. It’s part of a massive global market, long leveraged by the ultra-wealthy (Bezos, Gates, Rockefellers etc).

Contemporary and post-war art.

Masterworks lets you invest in multimillion-dollar artworks featuring legends like Banksy, Basquiat, and Picasso—without needing millions. Over 70,000 members have together invested more than $1.2 billion across over 500 artworks. So far, 25 sales have delivered net annualized returns like 14.6%, 17.6%, and 17.8%.*

Want access?

[Click here to skip the waitlist →]

_{Investing involves risk. Past performance not indicative of future returns. Reg A disclosures at masterworks.com/cd}