Do I have it right the robots are now plagiarizing each other and its somehow tanking the economy
Reposted from Anil Dash
DeepSeek made orders-of-magnitude leaps in all of these areas. (That means it was 10x to 100x more efficient.) Instead of gathering all that costly data, it _generated_ much of the necessary data — often using prior LLMs like Llama (Meta’s open source tool) and Qwen (an open source Chinese tool).

Comments