Models & Labs
China Challenges US AI Dominance with New Models
The Verge AIhigh confidence
Why it matters
- →China's AI models being open-source could democratize access and innovation globally.
- →The competition may drive faster advancements in AI technology.
- →US companies might need to reconsider their proprietary strategies to maintain a competitive edge.
China's AI sector is intensifying competition with the US as Moonshot and Alibaba introduce new models that claim to rival leading American systems. Moonshot's Kimi K3 and Alibaba's Qwen3.8 are both large-scale models, with parameter counts of 2.8 trillion and 2.4 trillion respectively, and are set to be open-source. This approach contrasts with the more guarded strategies of US companies, potentially altering the global AI landscape. The releases highlight the narrowing gap in AI capabilities between the two nations.
Read originalMore from The Verge AI
© The Verge AIMarket & Regulationmusic
Sony sues Udio over 30,000 AI-generated songs
Sony Music Entertainment has intensified its legal confrontation with Udio by filing a lawsuit over the alleged unauthorized use of more than 30,000 songs by the AI music generator. This action follows an earlier lawsuit that was restricted to 333 works due to a judicial ruling. Sony asserts that Udio's AI models were developed using a diverse collection of sound recordings, including those sourced from YouTube. In contrast, Universal Music Group and Warner Music Group have resolved their disputes with Udio and are now working together with the company. Sony aims to prevent further unauthorized use and seeks damages of up to $150,000 for each infringed work. This case exemplifies the ongoing struggle to reconcile AI advancements with existing music copyright frameworks.
The Verge AI
© The Verge AIVideo & Creative AIimage
Adobe's Indigo App Adds Generative AI Features
Adobe's Indigo camera app, initially designed for a natural SLR-like photography experience, is now embracing generative AI with its new AI Playground update. This update introduces features like generative filter effects, object removal, and chatbot feedback, enhancing the app's editing capabilities. Interestingly, Adobe is using Google's Nano Banana AI model rather than its own Firefly models for this experiment. While the update marks a shift from Indigo's original focus, it opens up new creative possibilities for users, albeit with ongoing considerations around AI transparency and responsible use.
The Verge AI
© The Verge AIVideo & Creative AImusic
AI Music Surprises with 'Semiramis' Dream'
1010Benja's track 'Semiramis' Dream' offers a fresh perspective on AI-generated music, proving it can be more than just algorithmic output. By leveraging Suno's generative AI, Benja has crafted a song that captivates with its dynamic jungle beat and his own compelling vocals. The AI's role is cleverly woven into the fabric of the music, making it difficult to distinguish from human elements. This track challenges the stereotype of AI music as monotonous, showing that when used creatively, AI can complement and elevate artistic expression. Benja's approach demonstrates that AI tools can be harnessed to push the boundaries of music production, rather than simply replace human creativity.
The Verge AI
More in Models & Labs
Models & Labsmodels
llama.cpp b10069 Release Enhances OpenCL Support
The b10069 release of llama.cpp brings notable improvements to OpenCL support, particularly targeting Adreno GPUs. By enabling broadcast for Adreno MUL_MAT and respecting view offsets, this update aims to boost performance for multi-stream operations on llama-server. The release also extends general GEMM/GEMV support for broadcast, which could optimize operations across different hardware setups. Although there are no revolutionary new features, these updates represent a consistent enhancement in compatibility and performance, especially for developers working with a range of hardware configurations.
llama.cpp Releases
Models & Labsmodels
llama.cpp b10075 release expands platform support
The b10075 release of llama.cpp marks a significant step in enhancing its compatibility across diverse hardware setups. With the addition of ROCm 7.2 support on Ubuntu, AMD GPU users can now enjoy improved performance. Windows users benefit from the inclusion of CUDA 13.3, ensuring better integration with NVIDIA GPUs. The update also brings Vulkan support, which optimizes GPU utilization for developers. Although no new model architectures are introduced, this release reinforces llama.cpp's role as a flexible and adaptable inference runtime for developers working in varied environments.
llama.cpp Releases
© TechCrunch AIModels & Labsmodels
Google Developing New AI Chip for Gemini Models
Google is reportedly working on a new AI chip, dubbed 'Frozen v2', aimed at significantly enhancing the efficiency of its Gemini models. Expected to be released by 2028, this chip could be six to ten times more efficient than current AI chips, potentially transforming Google's AI capabilities. This move aligns with a broader industry trend where tech giants are developing custom chips to reduce reliance on Nvidia and address AI computing capacity shortages. The anticipation of this chip has already positively impacted Google's stock, reflecting investor confidence in the company's strategic direction.
TechCrunch AI