China's first open-source 3D model released

Dear Readers,

Have you ever created a walkable 3D world – with just one sentence? HunyuanWorld 1.0 from China marks the beginning of a new era: text becomes space. What was once the preserve of game designers and VR studios is now available to everyone – open source, compatible with Unity, Unreal, and Blender. This is not only changing games, but also simulations, architecture, and education. It's a technological tipping point and a geopolitical signal: China is catching up. Fast.

In this issue, we explore the deeper implications: What does this opening mean for global markets, investment flows, and future dominance in the world of AI? Plus: A highly relevant US-EU tariff deal as an anchor of stability in transatlantic trade. And new figures show that AI creates growth, not job losses – provided that education and competition policy keep pace. A newsletter full of turning points – it's worth taking a second look.

In Today’s Issue:

All the best,

One sentence – and you're standing in the middle of your very own walkable 3D world: This is how the new era of generative AI begins with Tencent's HunyuanWorld 1.0, the world's first open-source model that creates immersive 360° scenes from text or a single image.

The model first generates a panoramic image (“world proxy”) and automatically creates a consistent, three-dimensional scene from it – including objects, depth, and editable mesh structures. This not only allows spectacular environments to be created for games or VR, but also reimagines real-world applications such as simulation training, architecture, and storytelling.

This is a milestone for the AI community: the combination of minimal prompts and maximum spatial impact dramatically lowers the barrier to entry. What used to take days of level design or scene building can now be done in minutes, leaving you ready to build, play, and explore in an interactive world.

Why it matters: It matters because it radically lowers the barrier to entry for 3D world creation using just text or a single image. It also matters because it empowers developers and creators worldwide through open-source access to next-gen scene generation tools.

Sources:

🔗 https://3d-models.hunyuan.tencent.com/world/

🔗 https://3d.hunyuan.tencent.com/sceneTo3D

🔗 https://github.com/Tencent-Hunyuan/HunyuanWorld-1.0

🔗 https://huggingface.co/tencent/HunyuanWorld-1

In 2 years you will be working for AI

Or an AI will be working for you

Here's how you can future-proof yourself:

Join 1,000,000+ pros at companies like Google, Meta, and Amazon that are using AI to get ahead.

Sign up and start learning AI

InternLM has released Intern-S1, a powerful new open-source multimodal model that leverages a 235B MoE architecture and a massive scientific dataset to achieve state-of-the-art performance on complex reasoning tasks, rivaling top closed-source competitors.

China and its open source models are catching up – and are now only slightly behind the top closed source models!

At their meeting on July 27 in Turnberry, Scotland, the US and the EU agreed on a framework agreement: in future, most EU imports will be subject to 15% US customs duties instead of the previously threatened 30%. In return, the EU will guarantee up to USD 600 billion in US investments and USD 750 billion in energy purchases, among other things. Exceptions remain: steel and aluminum will continue to be subject to 50% tariffs. The deal prevents escalation, brings market certainty, and signals a new standard in transatlantic trade policy.

Thanks to a wave of CapEx investments in data centers, chips, and energy infrastructure, AI has been making a significant contribution to US GDP since the beginning of 2024: in Q2/2025, >⅓ of growth (~2.5% annualized) was AI-driven.

This marks a structural turning point: from software-centric investments to “reindustrialization” through hard infrastructure. Markets and monetary policy are feeling the effects: robust demand for infrastructure financing, rising energy prices, higher lending volumes – and political pressure to promote AI infrastructure strategies.

Contrary to many fears, investment in AI does at the moment at least not lead to higher unemployment. Instead, companies are reporting more growth, product innovation, and employment – especially in knowledge-intensive areas. However, the gap is widening: higher-skilled workers are preferred, while industry concentration is increasing. For economists and policymakers, this means that education policy, retraining, and competition policy are needed to regulate imbalances and market power.

Working on the future of finance through AI? From trading algorithms to credit modeling and fintech infrastructure, we’re looking for work that explores the intersection of AI and financial systems.

Submit your paper or project to Superintelligence, the top AI newsletter with 200k+ readers, by emailing [email protected] with the subject line “Finance Submission”. We’ll contact you if we’d like to feature it.

The models presumably published as GPT-5 models on LMArena have disappeared again. Initial tests have achieved outstanding results. These are believed to be versions of GPT-5 full, mini, Nano, and OpenAI's open-source model, which is expected to be released this week.