DeepSeek gets Silicon Valley talking
Since the Chinese AI company Deepseek It has released an open version of its Thinking R1 model At the beginning of this week, many in the tech industry made big statements about what the company has achieved, and what it means for the state of artificial intelligence.
Venture capitalist Marc Andreessen, for example, to publish That Deepseek is “one of the most amazing and impressive hacks I have ever seen.”
The R1 appears to match the O1 Openai model on certain AI benchmarks. The company calls one of its models It cost only $5.6 million For training, compared to the hundreds of millions of dollars that American companies pay for their training.
He also appears to have achieved that in the face of US sanctions banning the sale of advanced chips to Chinese companies. He writes the MIT Technology Review The company’s success demonstrates how sanctions “drive startups like Deepseek to innovate in ways that prioritize efficiency, resource pooling, and collaboration.” (On the other hand, The Wall Street Journal reports DebSEC’s Liang Wenfeng recently told the Chinese premier that US export restrictions remain a bottleneck.)
Curai CEO Neal Khosla Give a simpler explanationclaiming that the company is “PSYOP CCP State” which was “faking low costs to justify price controls in hopes everyone will switch to them.” [to] Damage to AI competitiveness in the United States. (A community note was attached to his post noting that Khosla provides no evidence of this, and that his father Vinod is an Openai investor.)
Meanwhile, Journalist Holger Schaibitz suggested Deepseek “could represent the biggest threat to US stock markets – if a Chinese company can build a cutting-edge model at low cost, without access to advanced chips, then that will be poured into the industry.”
In response, Y Combinator CEO Garry Tan argued Deepseek’s success will actually be good for its American rivals. “If training models become cheaper and easier,” he wrote on
And head of Amnesty International in Meta Yann Lacon Argue against Looking at the DEPSEC announcement through the lens of China versus the United States. Instead, he suggested, the real lesson is that “open source models transcend proprietary models.”
“Deepseek has benefited from open research and open source (like Pytorch and Meta’s Llama),” Leccon wrote on LinkedIn this week. They come up with new ideas and keep others on top of their work. Since their work is published and open source, everyone can benefit from it. “
All the discussion seems to push consumers to try the product. As of Sunday afternoon, DEPSEC AI Assistant It is the best free app in the Apple App Store, ahead of ChatGpt.
