btw, DeepSeek first does its 'thinking' in Chinese, a fundamentally more efficient language apparently, am told

btw, someone asked DeepSeek whether AI shall replace humans, and as DeepSeek 'thinks' in the Chinese language, a more efficient language, and then translates the result to English, it, meaning DeepSeek returned a response that is a Chinese poem of its own creation, on another level, above PhD level

Huawei adds DeepSeek-optimized inference support for its Ascend AI GPUs

On January 27, the same day Nvidia’s stock price plummeted after the market fully appreciated what the Chinese LLM meant for the industry, China-based Huawei posted an article announcing that the distilled R1 AI model was then available for free via its ModelArts Studio platform. The tech company explicitly said this version is “Ascend-adapted,” as in Huawei’s Ascend data center GPUs.

Huawei chips are powering DeepSeek AI

I feel this should be a much bigger story: DeepSeek has trained on Nvidia H800 but is running inference on the new home Chinese chips made by Huawei, the 910C. pic.twitter.com/6IAgQlQ3ou

— Alexander Doria (@Dorialexander) January 28, 2025

... therefore not an inference but straight up, DeepSeek already ported over to ... drum roll ... Huawei Huawei Huawei, easy to do given original programming done in machine language (0s and 1s) and working all the time in collaboration

at 15:21 mark, the tell tale

Need this guy

Need $30

Discovery, that DeepSeek has a sense of humour, but in Chinese language, the VOLG as well as DeepSeek sense of humour