Our latest SCSP Substack discusses the implications of DeepSeek’s R-1, a new large language model (LLM) developed by a PRC AI firm. The model has matched the performance of leading U.S. closed models at much lower compute costs, challenging the U.S.'s assumed technological leadership.
Check out their conversation on the SCSP YouTube channel!
Memos to the President Podcast is a product of the Special Competitive Studies Project.
