Deep Seek And Destroy - Part Eight

The Prophecy Arrives In America

The Deep Seek AI bomb finally hit Wall Street yesterday, and now everyone knows who Deep Seek is.

Stocks like Nvidia, which makes the graphical processor units for AI, are still sliding on the second day of the Deep Seek hangover. Nvidia’s had almost $600 Billion loss in valuation in one day - a Wall Street record.

I have been harping on how Deep Seek stole the training data from Open AI from the murder of Suchir Balaji since the December 19th, 2024 presentation in Silicon Valley. Now that Deep Seek is a one day international sensation, perhaps my story will get more traction now.

Meanwhile, Deep Seek’s CEO went from unknown to instant celebrity in China, meeting with the Premier and his retinue yesterday.

Liang Wenfeng, a hedge fund trader, became the CEO of Deep Seek AI.

Not far from where I presented the Deep Seek info in Silicon Valley on December 19th, billionaire Mark Zuckerberg and Meta is headquartered. Of course, Meta took no notice of the warning.

Just down the street, Sam Altman had his Y Combinator Silicon Valley incubator before he found Open AI, and the restaurant where Elon Musk wrote the first check to Sam Altman for AI is just a few more steps down the street from the Rosewood Hotel.

No Open AI or Grok employees attended the talk, but the Stanford medical records custodian did. He said Google had taken all of Stanford’s medical records without permission in one of their joint ventures in health care management.

I presented the case for Deep Seek’s impact on Silicon Valley again on January 3rd. My strategy was that Silicon Valley knew AI, and they would want to hear my story about Deep Seek murdering Suchir Balaji for the AI training data to leapfrog and their competitors.

I even wrote a seven-part series called “Deep Seek And Destroy” on this Substack.

But nothing got the attention of Silicon Valley like a one Trillion and a half dollar haircut in market losses with no end in sight yesterday. Now, people want to know how Deep Seek leapfrogged the competition.

https://nypost.com/2025/01/27/business/nvidia-shares-fall-12-as-chinese-ai-startup-deepseek-triggers-panic/?utm_campaign=iphone_nyp&utm_source=twitter&utm_medium=social

Silicon Valley venture capital firms and the squawk boxes are issuing recriminations, hurling insults, and threatening lawsuits this morning. Never have the darlings of Silicon Valley fallen out of favor so quickly in one day.

Nvidia has not fallen 17%, with no hint of any price resistance yet to prevent a lower plunge. All manner of theories have abounded from Chinese sneaking GPU chips into China to radical new coding methods that leave the American billionaires like Elon Musk, Jeff Bezos, Sam Altman, and Mark Zuckerberg in the dust. I was being very specific on January 3rd of who trained Open AI with Suchir Balaji, and how his training drive did the same for Deep Seek.

https://x.com/RealGeorgeWebb1/status/1875239388906762736

I also introduced you to Alexander Wang of Scale AI last year, and this year at Davos, Wang is doing his best to calm the AI training nerves of everyone. Wang promises that the CCP snuck 50,000 Nvidia GPUs into China to do the training of their AI system for Deep Seek.

Interestingly, no one in the press asked Alexander Wang where these processors were and how did Deep Seek get them. The Davos and WEF press just wrote it down, “50,000 secret GPUs in China”. That may explain the Deep Seek leapfrog over the competition, and it may not. That remains to be seen.

By contrast, the New York Times and AI reporter Cade Metz says Deep Seek only used 2,000 older “Hopper” Nvidia chips to leapfrog all of Silicon Valley’s billionaires and their AI platforms in a single bound.

https://www.nytimes.com/2025/01/23/technology/deepseek-china-ai-chips.html

Whether the CCP got 50,000 of the latest Blackwell H-100 GPUs or not, almost all of the AI chips are made in Tawain, just 82 miles from Mainland China. And this is a glaring National Security vulnerability.

https://www.bakerinstitute.org/research/silicon-hegemon-could-china-take-over-taiwans-semiconductor-industry-without-invading

I have covered the key players from Google like Eric Schmidt and the founders of Deep Mind, who are certain to be at the center of the dealmaking with Kai-Fu Li in China.

Return here to the Paid section to find out the details. We have said Suchir’s training drive was used to train Deep Seek for six weeks. Now the experts are agreeing.