How Did DeepSeek Build Its A.I. With Less Money?

Published On Feb 12, 2025, 5:02 AM

A Chinese start-up named DeepSeek has developed a new powerful AI system using significantly fewer chips than competitors, only 2,000 compared to the typical 16,000. This breakthrough has reduced its computing costs to about $6 million, a fraction of what major firms like Meta invest in similar technologies. The innovation could lead to cheaper and more accessible AI technology, potentially reshaping the market landscape.

Stock Forecasts

NVDA

Negative

DeepSeek's cost-effective AI solution could challenge existing industry leaders, providing a competitive edge through lower operational costs. This may affect companies that rely on traditional AI infrastructure, such as Nvidia and Meta, negatively impacting their stock prices due to potential loss of market dominance.

ARKQ

Positive

Investing in emerging technologies that focus on efficiency, such as DeepSeek, could yield high returns as they potentially disrupt the AI market. Companies embracing innovative AI solutions may see favorable growth.

Related News

Paris showcases the country’s multibillion-dollar bet on artificial intelligence, but the summit also highlights differences in culture, liberty and national priorities.

Non-tech investors interested in AI and its effect on the stock market need only look for one clear warning sign: any cuts in AI spending by the hyperscalers.

NVDA
MSFT

(Reuters) -A consortium led by Elon Musk said on Monday it has offered $97.4 billion to buy the nonprofit that controls OpenAI, months after the billionaire sued the artificial intelligence startup to block it from transitioning to a for-profit firm. Altman promptly posted on X: "no thank you but we will buy twitter for $9.74 billion if you want." Musk criticized a $500 billion OpenAI-led project called Stargate announced with great fanfare at the White House just after President Donald Trump returned to office, suggesting the investors involved lacked the funding for the project.

NVDA
AAPL