Introducing Nova Multimodal Embeddings! The first embeddings model under the Nova SOTA foundation model family and the first model to generate embeddings for multimodal inputs within the same embedding space....
Read more →Model customization is having a moment. If I use an evolution analogy, I see three phases so far: Transfer learning matured in vision (think the ImageNet era), then jumped to...
Read more →Anthropic's new Agent Skills make extending models lighter than spinning up full MCP servers — and they play nicely together. Instead of loading every tool, your agent loads just the...
Read more →We don't lose trust because models hallucinate — we lose it because our rules are ambiguous, outdated, or impossible to decide exhaustively. 1. Translation: Natural language → formal logic, with...
Read more →Bias in AI image generators isn't just an ethics issue — it's a brand, compliance, and revenue risk. Amazon's Nova Canvas demonstrated the lowest bias; it was also the only...
Read more →