AI Daily: The Latest AI Trends and Updates – July 28

This week at the 2025 World Artificial Intelligence Conference (WAIC), the AI landscape entered a new phase of maturity and multidimensional impact. From full-scale autonomous driving powered by world models to GPT-5 integrations in Copilot, from immersive 3D generation to major funding pushes in AGI, and even Amazon’s entrance into AI-powered wearables — the boundaries of what AI can do continue to expand. Here are the five most important stories you need to know.

Autonomous Driving Enters the World Model Era

Autonomous driving took center stage at WAIC 2025, where companies like MogoAuto and SenseTime’s Yueying unveiled their use of large physical world models. MogoAuto showcased its “vehicle-road collaboration” approach, using roadside sensors and communication infrastructure to collect real-world data to train its MogoMind model. This system can identify road conditions, offer driving suggestions, and is already deployed in several L4 autonomous vehicle fleets used for public transit and sanitation.

Commentary: World models mark a shift from simple perception to deep contextual understanding. With deployments already in real-world urban environments, AI-powered autonomy is moving from demo to reality.

Microsoft to Launch “Intelligent Mode” in Copilot with GPT-5

Microsoft has announced plans to launch an Intelligent Mode for Copilot this August, powered by the upcoming GPT-5. This new mode will intelligently route user queries to the most appropriate tools and adapt its responses in real time. Meanwhile, OpenAI is developing a model router to automatically distribute GPT-5 queries based on context, aiming to turn Copilot into a truly multi-modal, dynamic AI assistant.

Commentary: This is more than an upgrade — it's a paradigm shift. Intelligent routing will blur the lines between tools, models, and interfaces, redefining the user-AI interaction experience.

Tencent Launches Hunyuan 3D World Model 1.0

At WAIC, Tencent unveiled and open-sourced Hunyuan 3D World Model 1.0, the first Chinese-developed model capable of generating interactive, physics-aware 3D environments from a single sentence or image. Users can freely roam generated scenes in 360° immersive mode, edit them, and export standard mesh files for use in gaming or digital twin applications. Tencent also announced further open-source initiatives to accelerate the 3D AIGC ecosystem.

Commentary: Hunyuan 3D could be a game-changer for embodied intelligence and creative industries alike. As 3D content generation becomes more accessible, entire workflows in gaming, film, and virtual worlds could be redefined.

Anthropic Eyes $5B Funding Round, Targets $150B Valuation

According to the Financial Times, Anthropic is in talks to raise $3–5 billion, which would push its valuation to $150 billion. The funding is intended to support further development of AGI systems and expand the Claude model family. Investor interest remains high as Anthropic positions itself as a major contender in the race for safe, scalable artificial general intelligence.

Commentary: If completed, this round would place Anthropic among the top-valued AI companies globally — a clear sign that investor confidence in long-term AGI bets is still accelerating.

Amazon Acquires Bee AI: Enters AI-Powered Wearables Market

Amazon has acquired Bee AI, a San Francisco-based startup known for its $50 AI wearable that listens to conversations and provides AI-generated summaries and reminders. Bee’s device gained attention at CES earlier this year for offering to-do list creation, alerts, and personalized daily digests. It also supports a $19/month subscription and an Apple Watch companion app. The deal was confirmed by co-founder Maria de Lourdes Zollo via LinkedIn, though terms were undisclosed.

Commentary: Bee AI gives Amazon a head start in the wearable AI space. With Alexa integration likely on the horizon, this could signal Amazon’s strategy to extend intelligent assistants beyond the home and into everyday life.

Dreamface AI Update: Dream Avatar 3.0 Now Live

One More Thing: Dream Avatar 3.0 is officially live! Premium users can now experience one of the world’s most advanced full-body digital human technologies. Whether you're creating lifelike avatars for content, storytelling, or virtual presence, Dream Avatar 3.0 offers unprecedented realism, motion accuracy, and personalization. Try it now and bring your digital identity to life like never before.