Anthropic, the company behind AI model Claude, has filed its S-1 prospectus with the SEC, planning to list on the Nasdaq exchange under the ticker symbol 'ANT'. The company had an $80.7 billion net loss on $16.3 billion in revenue in the first half of this year, mainly due to accounting charges related to notes issued for funding. Anthropic has secured billions of dollars in deals for access to compute power, including $3.1 billion per month with SpaceX, $44.9 billion with Microsoft, at least $114.5 billion with Google, and $110 billion with Amazon. The company aims to 'revolutionize every area of human endeavor' with AI, but notes the risks and challenges associated with this technological revolution.

Cerebras Systems has deployed a trillion-parameter AI model on its CS-2 AI supercomputer. This model is called the Frontier model, and it has 1.2 trillion parameters, making it the largest AI model to date. The model is used for large language models and other applications. The deployment was made using Cerebras' Wafer-Scale Engine (WSE) chips, which provide a large number of processing cores and high-bandwidth memory. The WSE chips are integrated into the CS-2 system, which is specifically designed for AI workloads. The Frontier model was trained on the BERT and RoBERTa language models, and it achieved state-of-the-art results on several benchmarks.