Comment: “OpenAI’s model can be harnessed and utilized in ways that outperform DeepSeek in every way. Consider something as simple as a hammer ?, a tool that can build far beyond its own scale and size. These LLM’s will never go away but instead are generating guard rails and ecosystems for us to build within and or on top of.
“So what remains? Who am I most comfortable sharing my data with? Who is GDPR compliant, Fedramp compliant, allows your models to be your own, belongs to a country that has strong ethical compliance, rights for its citizens and a great quality of life? “
My Response: DeepSeek does not need to be as great as OpenAI to be successful. It simply needs its products to be above performance average as it has a massive cost competitiveness. In the Android smartphone market, China won, and controls most parts of the world, because it delivers products people can afford. So, this is not really about frontal competition, this is about breaking the economic veil that one must be so loaded with money to do AI.
That myth that one must raise $billions for core AI work has been broken. And with that, Nvidia will not recover easily because its “new GPU” thesis has been invalidated. DeepSeek may not be valued at up to $50 billion but it has knocked out more than $300 billion in global AI market caps.
On data sharing, no one eats data, and we overestimate metadata-level digital privacy in an era people share their lives on social media. That is the reason people are buying Chinese phones, Chinese EVs, etc even as the US and Europe block them, denying their citizens the best EVs in the world right now. The Chinese EVs are well ahead of what we have in the US today.
On GDPR, Fedramp compliant, rights, etc, those are European and Western inventions that most people do not really care about except when in their domains. Unless the US puts an executive order to ban DeepSeek, I expect Silicon Valley startups to join the DeepSeek open model framework to find a positioning since they do not have $billions for Nvidia anymore. When that happens, everyone will understand that people do not really care about those GDPRs. With GDPR, you pay $200m for what someone is giving you $6m with/out GDPR, your choice?
On quality of life, the irony is that the DeepSeek advancement is how to keep improving the quality of life for ALL Nations, not just a few. This is how Microsoft’s boss sees it: “To see the DeepSeek new model, it’s super impressive in terms of both how they have really effectively done an open-source model that does this inference-time compute, and is super-compute efficient. We should take the developments out of China very, very seriously”.
Remember: “Microsoft asked hundreds of China-based AI staff to relocate out of China by July 2024”. And that China has done this. The US must learn of the need for the rise of ALL, not just one region. China is likely going to win the AI development future in the Global South now.
That said, there is no need for panic because if China makes AI development easier and accessible, the global market will grow and everyone will benefit, assuming some presidents think that is a better outcome for the world.
