DeepSeek AI Disrupts Market, Challenging U.S. AI Giants

Chinese AI startup DeepSeek is making waves, with its AI assistant surpassing OpenAI’s ChatGPT on Apple’s U.S. App Store. Using Nvidia’s lower-capability H800 chips, DeepSeek claims to have trained its model for under $6 million, sparking concerns among U.S. AI firms and investors.

Top AI developers acknowledge DeepSeek’s technological advancements but remain skeptical of its full development costs. While OpenAI CEO Sam Altman praised DeepSeek R1’s efficiency, Nvidia emphasized the growing demand for its chips. Meanwhile, Snowflake added DeepSeek models to its AI marketplace following increased customer interest, despite initial concerns over Chinese AI integration.

U.S. AI labs are analyzing DeepSeek’s V3 model, which has accelerated China’s AI progress to just six months behind U.S. counterparts. However, skeptics argue the total cost of development likely exceeded $1 billion, as the $6 million figure only covered the final training phase.

DeepSeek’s success highlights the power of open-source AI, offering an alternative to proprietary models like ChatGPT. Venture capitalist Marc Andreessen called DeepSeek R1 “a profound gift to the world.” With U.S. tech giants reporting earnings soon, expectations around AI dominance are shifting, challenging the assumption that only billion-dollar firms can lead AI innovation.