Wedbush analyst Dan Ives has expressed skepticism over Chinese startup DeepSeek’s assertion of developing its low-cost AI model, R1, for under $6 million.
What Happened: DeepSeek has stated that it developed its R1 model in just two months with an investment of $5.6 million. This statement has raised eyebrows, considering the billions of dollars U.S. tech companies are investing in AI development.
For instance, OpenAI’s GPT-4 model cost over $100 million to train, and future models could cost upwards of $1 billion.
Ives referred to DeepSeek’s claim as “likely a fictional story.”
Don’t Miss:
OpenAI has also found evidence suggesting that DeepSeek may have used its proprietary models to train the open-source competitor.
Why It Matters: Earlier, Josh Kushner, a major investor in OpenAI through his venture firm Thrive Capital, criticized colleagues who publicly supported DeepSeek. He alleged that the Chinese model was trained using U.S. technology and possibly violated export controls.
Elon Musk also responded to a new clip with a blunt “obviously” after Scale AI’s Alexandr Wang told CNBC that DeepSeek had amassed 50,000 advanced chips that it couldn’t publicly acknowledge due to export controls.
Trending: Nancy Pelosi Invested $5 Million In An AI Company Last Year — Here’s How You Can Invest In Multiple Pre-IPO AI Startups With Just $1,000
Oculus founder Palmer Luckey has also dismissed the $5.6 million figure associated with DeepSeek as false, while billionaire investor Bill Ackman questioned whether DeepSeek’s hedge fund affiliate profited from Nvidia Corp.’s (NASDAQ:NVDA) stock plunge following the startup’s emergence.