Theara Coleman
Challenging the 'bigger is better' narrative
The U.S. has imposed sanctions on advanced chip sales to slow down progress in AI elsewhere. At the same time, U.S. tech giants rely on sprawling data centers powered by Nvidia GPU chips. These sanctions, first imposed under the Biden administration, have "cut China off from critical AI hardware, forcing its developers to innovate with far fewer resources," said The Spectator. This is the environment that led to the birth of DeepSeek. The relatively unknown Chinese AI startup has "emerged as a formidable challenger to the 'bigger is better' narrative" while achieving the seemingly impossible: "delivering performance comparable to the West's cutting-edge models" at a much lower price point.
In a 22-page paper that sent shockwaves through the tech world, DeepSeek revealed the workings of its new AI model called DeepSeek-R1. A team of researchers claimed to have used around 2,000 of Nvidia's H800 chips, drastically undercutting the number and cost of more advanced H100 chips typically used by the top AI companies. The chatbot run on the R1 model distinguishes itself from competitors like ChatGPT by "articulating its reasoning before delivering a response to a prompt," said Bloomberg. DeepSeek said its R1 release performs on par with the latest iteration of ChatGPT. The company also offers licenses for developers interested in creating chatbots with the technology "at a price well below what OpenAI charges for similar access." The efficiency and cost-effectiveness of the model "puts into question the need for vast expenditures of capital to acquire the latest and most powerful AI accelerators from the likes of Nvidia," Bloomberg added. It also highlights "U.S. export curbs of such advanced semiconductors to China," which were supposed to "prevent a breakthrough of the sort that DeepSeek appears to represent."
No comments:
Post a Comment