In this podcast episode, our AI friends discuss these recent AI developments:
Anthropic's Claude AI now has enhanced style adaptation capabilities allowing it to better match different writing styles. Users can provide a sample of text and Claude will analyse it to replicate the writing patterns in its responses. Alongside this, Anthropic has also introduced three preset writing modes: formal, concise, and explanatory.
Luma AI's Dream Machine platform offers AI-powered video creation with intuitive interfaces and advanced features like maintaining character consistency across videos generated from a single image. The platform is subscription-based and aims to be user-friendly for both casual creators and professionals. Dream Machine utilises a new image generation foundation model, Luma Photon, which is built on Luma's Universal Transformer architecture, making it eight times faster and more cost-efficient than comparable models.
Researchers at MIT have developed an AI tool that generates realistic satellite images depicting potential future flooding scenarios. The method combines a generative AI model with a physics-based flood model to create accurate visualisations. The team tested the tool by generating images of Houston after a storm similar to Hurricane Harvey and found it produced more realistic results than an AI-only method.
Anthropic has also introduced the Model Context Protocol (MCP), an open-source standard for connecting AI assistants to data sources. MCP enables developers to create two-way connections between data sources and AI-powered applications, allowing AI models to access and utilise information from various systems.
Nvidia has unveiled Fugatto, a new AI audio model capable of synthesising novel sounds by combining existing audio traits. Fugatto can adjust distinct audio traits, combine traits to generate new sounds, and perform tasks like changing emotion in spoken text and isolating vocal tracks
Share this post