A new project called Swiftlet enables running large language models on Apple devices with minimal resources. The tool allows an 80B parameter Qwen model on Mac and a 35B variant on iPhone.
Swiftlet, shared on GitHub, demonstrates efficient model compression and optimization for Apple silicon. The project achieves these results through quantization and memory optimization techniques, making large language models practical for consumer hardware.
The ability to run 80B parameters in 4.3GB of RAM represents a significant step toward on-device AI inference. iPhone users can access a 35B model, enabling local language processing without cloud connectivity or API dependencies.
The project generated interest on Hacker News with 130 points and 46 comments, with developers discussing potential applications and technical approaches. This advancement aligns with broader industry trends toward edge computing and reducing reliance on centralized AI services.
Swiftlet is available as open source on GitHub, allowing developers to experiment with local inference on Apple devices.
The U.S. Department of Energy has announced the Genesis Open Models Initiative, a program aimed at developing and democratizing artificial intelligence models for scientific research and industrial applications.
Artificial intelligence tools prove insufficient for protecting online communities from AI-generated harms. Human moderators remain essential for effective content oversight.
Rippling unveiled AI Spend Console this week, a tool that monitors individual and team AI spending after the HR software company burned through millions on AI in recent months.
Spelman College President Dr. Ayanna Howard discussed federal funding rollbacks affecting HBCUs and artificial intelligence's influence on college graduates in a recent interview.