⚡ Healthcare breakthrough
Plus: 930% profit surge, new app, datasets, engineering insights and so much more
Here's everything you need to know in Generative AI today:
INNOVATION & RESEARCH
The latest, pivotal innovation and research in Generative AI
➲ Google introduces Med-Gemini, a suite of AI models fine-tuned for healthcare, surpassing GPT-4 and other models on 10 out of 14 medical benchmarks. It achieves a 91.1% accuracy on the MedQA benchmark (medical question-answering ability), outperforming the previous best by 4.6%. Explore >
➲ Anthropic launches Claude as an iPhone iOS app. Claude joins OpenAI's ChatGPT and Google's Gemini with apps available to both free and paid users. Explore >
➲ The mysterious ‘gpt2-chatbot’ vanished from the Lmsys arena after a one-day appearance, sparking widespread speculation in the AI community. Explore >
INDUSTRY & SOCIETY
The latest impacts of Generative AI on industry and society
➲ Eight major U.S. newspapers sue OpenAI and Microsoft for allegedly using their copyrighted articles to train AI chatbots like ChatGPT and Copilot. The NY Daily News, The Chicago Tribune, among others, are included in the filing. Explore >
➲ Samsung's Q1 2024 profit surged by 930% in Q1 of 2024, driven by high demand for its AI memory chips and servers. Explore >
➲ University of Pennsylvania becomes the first Ivy League university to offer a Master’s Degree in AI, with courses starting in Spring 2025. Explore >
ENGINEERING SIMPLIFIED
Fresh insights for you or your Generative AI Architects & Engineers
➲ A new list of high-quality datasets, the most valuable asset in LLM development, and tools for LLM fine-tuning is released. Explore >
➲ XTuner an efficient, flexible versatile toolkit for large model fine-tuning, supporting continuous pre-training, instruction fine-tuning, and agent fine-tuning with QLoRA, LoRA, and full parameter training algorithms. Explore >
➲ Leverage PyTorch's in-place operations for memory-efficient indexing in complex data manipulations. This technique directly manipulates tensor data without the need for intermediate tensors, reducing memory usage and enhancing computation speed, especially for large datasets or models. Explore >
Lead the curve | every day of the week with The Generative AI Briefing