Ever since the internet, many companies have been disrupted from the long standing ways of doing things and as long as does not affect you personally most of it we can live with. Some industries were seen as harder to disrupt and narratives of why they are more stable lead to investments of billions of dollars. One of these groups was the AI use of computer chips to generate AI. The leader was Nvidia chips which have a high margin and Chat GPT which cost at least $100 million. Then something changed.

In an article by Meaghan Tobin, Paul Mozur and Alexandra Stevenson of the New York Times News Service, the change was a Chinese company with an AI app called DeepSeek.

DeepSeek’s origins are in finance, not technology for technology’s sake. Its parent is a Chinese hedge fund called High Flyer. It was using AI to make bets in the Chinese stock market.

In the Chinese stock market, if retail investors are jumping in and out of stocks, the Chinese government will mandate speculation is to be kept at a minimum for the long-term viability of the market. High Flyer decided to align itself better with the Chinese government priorities: advance AI.

High Flyer CEO Lu Zhengzhe said we want to do things with greater value and things that go beyond the investment industry. High Flyer started a startup called DeepSeek.

DeepSeek’s latest model for AI is believed to be nearly as powerful as American rivals but far more efficient. Its success is despite Washington efforts to limit Chinese access to the advanced chips needed for AI.

DeepSeek’s revenue model did not rely on making consumer-facing AI products for revenue, and only this month released its first chatbot, which allows anyone to generate text and photos with simple commands. High Flyer has been subsidizing DeepSeek with profits from AI trading.

The research and development approach allowed DeepSeek to side-step stringent regulations the Chinese government has placed on AI use by the public.

DeepSeek is run by its CEO Liang Wenfeng, an engineer who studied at Zhejiang University at Hangzhou.

Unlike many Chinese companies, which tend to focus on hiring programmers. Mr. Liang has given a reputation for employing people from outside of computing. Poets and humanities majors from top Chinese universities train the model to write classical Chinese poetry and ace questions from the country’s difficult college entrance examination. Most of the staff graduated from top universities in China – they are very smart and very young.

DeepSeek uses 10,000 advanced Nvidia A100 chips.

The DeepSeek model allows developers to build applications using its model and at the moment it was one of the least expensive models to work with. (Think of the Apple ecosystem – Apple is the foundation and developers build apps from the foundation).

Linking to dividend paying stocks, it seems every industry has the ability to be disrupted. You may believe that your SWOT analysis means few threats, but the lesson is for now. Things can change.

There are more questions than answers, till the next time – to raising questions.