the-decoder3d
Nvidia Blackwell doubles AI training performance in early benchmarks
In the MLPerf Training 4.1 benchmarks, the Nvidia Blackwell platform delivered 2.2 times more performance per GPU compared to Hopper in the LLM benchmark Llama 2 70B fine-tuning and 2 times more ...
the-decoder3d
Generative AI reduces demand for some freelance jobs in writing, coding and design, study says
The November 2022 launch of ChatGPT marks a turning point: demand for "automation-prone jobs" decreased significantly, while more manual jobs proved more resistant to change. | Image: Zhu et al.
the-decoder3d
German music rights group GEMA sues OpenAI over unlicensed song lyrics in ChatGPT
German music rights organization GEMA filed a test case against OpenAI in Munich Regional Court on November 13, making it the first collecting society worldwide to take legal action against an AI ...
the-decoder4d
MIT's test-time training makes AI language models better at solving hard puzzles
MIT researchers have found a way to significantly improve how AI language models solve problems using a technique called test-time training (TTT). The team has set a new record on a challenging AI ...
the-decoder6d
OpenAI's new "Orion" model reportedly shows small gains over GPT-4
The Information reports that OpenAI's next major language model, codenamed "Orion," delivers much smaller performance gains than expected. The quality improvement between GPT-4 and Orion is notably ...
the-decoder4d
Anthropic delays next-gen AI model Opus 3.5 with no new timeline
Anthropic CEO Dario Amodei has confirmed that the company is working on a new version of its flagship AI model, Claude Opus, but he's staying tight-lipped about when it will actually drop. The update, ...
the-decoder4d
Baidu unveils LLM smart glasses and new image generator
At the conference, Baidu CEO Robin Li introduced I-RAG, a text-to-image system that aims to reduce inaccuracies in AI-generated images where the output doesn't match the text input or contains ...
the-decoder6d
Google's ReCapture lets users change camera angles in existing videos
Google researchers have developed ReCapture, a new AI technique that allows users to modify camera movements in videos after they've been recorded. The system aims to bring professional-grade video ...
the-decoder4d
Alibaba's Qwen AI unit releases capable new models to help developers write and analyze code
Called Qwen-2.5-Coder, these models help developers write, analyze, and understand code. The new series includes six different model sizes, ranging from 0.5 to 32 billion parameters, to accommodate ...
the-decoder5d
OpenAI co-founder Sutskever predicts a new AI "age of discovery" as LLM scaling hits a wall
Leading AI companies are changing course. Instead of developing ever-larger language models, they are focusing on test-time compute, which uses more processing power during model execution rather than ...
the-decoder5d
Google DeepMind releases AlphaFold 3 for scientific use
Google DeepMind has released AlphaFold 3 on GitHub. The release includes both code and model parameters. According to AlphaFold 3's terms of use for the model ...
the-decoder5d
AI benchmark FrontierMath exposes the relativity of measuring artificial intelligence
A new benchmark called FrontierMath, created by leading mathematicians, shows that current AI systems perform poorly on complex mathematical problems, despite high scores on simpler tests. According ...