In January 2025, Hangzhou DeepSeek Artificial Intelligence Basic Technology Research Co. Ltd. launched a breakthrough large language model (LLM) on the generative artificial intelligence (GenAI) landscape. DeepSeek was a disruptive application: it had modest hardware requirements and low training costs and was released as a modified open-source app, allowing users access to the weights and codes DeepSeek used in its learning process. Users could run the app locally, reducing their reliance on application program interfaces (APIs) and the associated costs. The financial markets responded quickly and affirmatively to this innovation and to the implications for the AI value chain. But competitors were caught by surprise and threatened. What were the advantages and disadvantages of their positioning in light of DeepSeek’s emergence and how could they compete with the disruptor? At the same time, how could DeepSeek transform its early-stage venture’s rapid evolution into an efficient scalable solution?
看看哪些人也有訂購？