anthropic Anthropic Emerges as World's Most Valuable AI Safety Startup Founded by former OpenAI employees, Anthropic develops the Claude AI models with a strong focus on safety. The San Francisco-based company now holds a staggering valuation of nearly $1 trillion.
anthropic Anthropic Emerges as World's Most Valuable Pure-Play AI Company Founded by former OpenAI employees, Anthropic is an American AI safety company known for developing the Claude line of large language models. The San Francisco-based startup currently holds a staggering estimated valuation of $965 billion.
nlp New Survey Explores How Large Language Models Transform NLP Tasks A comprehensive new survey examines the impact of large pre-trained transformer models like BERT on natural language processing. The paper highlights methods like fine-tuning and prompting while addressing current limitations.
natural language processing Survey Explores How Large Language Models Transform NLP Tasks A new survey examines the profound impact of large, pre-trained language models like BERT on the field of natural language processing. The paper categorizes recent advancements into fine-tuning, prompting, and text generation methods.
microsoft Microsoft and NVIDIA Unveil 530-Billion Parameter Megatron-Turing Language Model Microsoft and NVIDIA collaborate to build MT-NLG, a 530-billion parameter generative language model that sets new accuracy standards across various natural language tasks.
stanford Stanford Researchers Analyze Risks and Opportunities of AI Foundation Models Over 100 Stanford researchers publish a massive paper examining how large-scale foundation models like GPT-3 are shifting the AI paradigm. The study highlights the benefits of AI homogenization while warning of inherited flaws and poor interpretability.
microsoft Microsoft Advances AI at Scale with DeepSpeed Compression and Z-Code Models Microsoft Research highlights major progress in its AI at Scale initiative, showcasing new tools for model compression and multilingual translation improvements. These innovations aim to reduce the latency and cost barriers of running massive AI models.
stanford Stanford Researchers Examine AI Paradigm Shift Driven by Foundation Models Over 100 Stanford researchers publish a massive 200-page study analyzing the profound impact, emergent capabilities, and inherent risks of large-scale pretrained foundation models like GPT-3.
jurassic-1 AI21 Labs Unveils Jurassic-1, a Massive Language Model to Rival GPT-3 AI21 Labs introduces Jurassic-1 Jumbo, a 178-billion-parameter language model designed to compete directly with OpenAI's GPT-3. The model demonstrates impressive capabilities by generating historically accurate rap lyrics without explicit prompting.
codex OpenAI Codex Tackles Python Generation Using Repeated Sampling A new GPT language model called Codex demonstrates strong Python writing capabilities by generating solutions from docstrings. The model achieves a high success rate through a strategy of repeated sampling.
artificial intelligence Global Scientists Unite to Study Risks of Large Language Models As tech giants like Google and OpenAI rapidly deploy large language models into consumer products, a massive collaborative project called BigScience aims to study the ethical and environmental dangers of this technology before it causes widespread harm.
wu dao China Unveils Wu Dao, a Massive AI Model Ten Times Larger Than GPT-3 The Beijing Academy of Artificial Intelligence introduces Wu Dao, a multi-modal AI system trained on 1.75 trillion parameters. Despite its massive scale and creative capabilities, experts doubt brute-force deep learning alone leads to general artificial intelligence.
eu ai act EU AI Act Faces Uncertainty Over Innovation and Safety Balance The European Union's upcoming AI Act applies a product safety framework to artificial intelligence, but critics worry this approach struggles with general-purpose models like ChatGPT. It remains unclear whether the law encourages responsible AI development or stifles technological progress.
google brain Google Releases Open-Source Switch Transformer with 1.6 Trillion Parameters Google Brain open-sources the Switch Transformer, a massive 1.6-trillion-parameter AI language model that trains up to seven times faster than its predecessor. The model achieves this speed by using a mixture-of-experts approach to keep computational costs low.
artificial intelligence Breaking Down ChatGPT: How AI Chatbots Actually Process Language A new visual guide reveals the complex math and systems powering large language models like ChatGPT. The explainer also provides a helpful glossary for understanding the rapidly evolving AI industry.