Alibaba's Qwen team releases QwQ-32B-Preview, a 32.5 billion-parameter AI model that outperforms OpenAI's o1 models in certain benchmarks, available for download under a permissive license.
Anthropic releases Claude 3.5 Haiku, a new AI model that outperforms its predecessor in specific benchmarks, offering improved coding recommendations, data extraction, and content moderation.
The ARC-AGI benchmark, a prominent test for artificial general intelligence, is closer to being solved, but its creators argue that this progress highlights the test's design flaws rather than a significant research breakthrough.
Eugenio Marongiu, an alpha tester on OpenAI's Sora, creates stunning 'Surreal Elderhood' piece, showcasing the potential of text-to-video generation
AWS updates Amazon Bedrock with features to streamline app testing, including RAG evaluation, custom connectors, and reranking models for improved performance and cost-effectiveness.
AWS announces updates to SageMaker HyperPod, focusing on efficient model training and fine-tuning, with features like flexible training plans and resource pooling.
AWS introduces a managed ecosystem of AI apps within SageMaker, streamlining ML workflows and addressing security concerns
Amazon Web Services launches new tools to improve AI model reliability, including Automated Reasoning checks, Model Distillation, and multi-agent collaboration features
AWS announces SageMaker Unified Studio, a single platform for data discovery, preparation, and AI model development, along with two new additions to its SageMaker product family.
Copyright © 2024 Starfolk. All rights reserved.