Chinese AI company DeepSeek has made waves in the tech industry after releasing an open version of its reasoning model, R1, earlier this week. The announcement has stirred significant conversations about its implications for the future of artificial intelligence, with some calling it a breakthrough and others expressing scepticism.
Industry reactions highlight mixed views
Prominent voices in the tech world have shared their thoughts on DeepSeekโs achievement. Venture capitalist Marc Andreessen described it as โone of the most amazing and impressive breakthroughs Iโve ever seen.โ DeepSeekโs R1 model reportedly rivals and sometimes surpasses OpenAIโs o1 model in specific AI benchmarks.
Deepseek R1 is one of the most amazing and impressive breakthroughs Iโve ever seen โ and as open source, a profound gift to the world. ๐ค๐ซก
— Marc Andreessen ๐บ๐ธ (@pmarca) January 24, 2025
Even more striking is DeepSeekโs claim that one of its models was trained for just US$5.6 millionโfar less than the hundreds of millions typically spent by leading US companies.
This accomplishment comes despite US sanctions restricting Chinese companies from purchasing advanced chips. According to the MIT Technology Review, these restrictions have pushed companies like DeepSeek to innovate by prioritising efficiency and collaboration. However, The Wall Street Journal notes that DeepSeekโs founder, Liang Wenfeng, has acknowledged that export restrictions remain a bottleneck for the company.
deepseek is a ccp state psyop + economic warfare to make american ai unprofitable
— Neal Khosla (@nealkhosla) January 24, 2025
they are faking the cost was low to justify setting price low and hoping everyone switches to it damage AI competitiveness in the us
dont take the bait
Not everyone is convinced of the companyโs transparency. In a post, Neal Khosla, CEO of Curai, claimed that DeepSeek could be a โCCP state psyop,โ alleging that the company might be misleading about its costs to gain a competitive edge. However, a Community Note attached to his post highlights the lack of evidence for his claim and points out that his father, Vinod Khosla, is an investor in OpenAI.
Implications for the tech industry
DeepSeekโs potential to produce cutting-edge AI models at a low cost has raised questions about the impact on the global AI market. Journalist Holger Zschaepitz suggested that this development could threaten the value of billions of dollars invested in the US AI industry. However, Y Combinator CEO Garry Tan offered a more optimistic view, arguing that cheaper and faster model training could increase the demand for AI applications and boost the industry overall.
Chinaโs #DeepSeek could represent the biggest threat to US equity markets as the company seems to have built a groundbreaking AI model at an extremely low price and w/o having access to cutting-edge chips, calling into question the utility of the hundreds of billions worth ofโฆ pic.twitter.com/wMVyczpRgT
— Holger Zschaepitz (@Schuldensuehner) January 24, 2025
Metaโs Chief AI Scientist, Yann LeCun, shifted the conversation away from a US-versus-China narrative. He emphasised the significance of open-source AI models, pointing out that DeepSeek built upon open research tools like PyTorch and Llama. LeCun argued that open-source innovation benefits the AI community, including competitors.
Consumers embrace DeepSeek
While the debate rages on among industry experts, consumers have shown great interest in DeepSeekโs AI assistant. As of Sunday afternoon, it became the top free app in the Apple App Store, surpassing ChatGPT in popularity. This surge reflects growing curiosity and excitement about DeepSeekโs potential to shake up the AI landscape.