It's time to change the US education policies and I hope Donald Trump and Elon Musk can take this on!

A not-so-secret confession: as a serial entrepreneur, I am great at managing costs and figuring out how to do things on a shoestring. When the news broke this morning of DeepSeek building out a free top-rated AI app, the US markets went into a tailspin and left many investors in disbelief. President Trump called the occurrence “a wake-up call”. The first thing on my mind was “I told ya so”.

Okay, this may sound arrogant and entitled, but I literally wrote a 40-page research this past Fall about how to use mathematics to reduce machine learning computing operations. The answer: there are basic mathematics that AI algorithm designers often ignore. These simple techniques dramatically optimize AI and eliminate the need for massive computing power, even in computing-hungry models like ChatGPT.

Look, machine learning works. If you task a computer with finding an optimal way, the computer will eventually find a way. If you provide the same computer with unlimited computing power, the computer will find that optimal way faster. Well-funded US companies like OpenAI, Uber and, of course, Amazon have always pumped cash into unlimited data servers to process data. However, why waste investor cash on chips when math does the trick?

The answers are many. A requirement for many chips can convince investors that a given AI platform has a natural barrier to entry: a difficult-to-replicate business model. The availability of extreme computing power also helps companies process a lot of data with fairly rudimentary machine learning skills. Chips companies are great at marketing their products. The list goes on.

So why did the Chinese company just seemingly blow the US AI industry out of the water? My answer is Math. As I have said for many years, US Math education is in dire crisis. At the same time, Chinese math is infinitely superior. I have had a great privilege of teaching applied mathematics at Cornell University for the past eight years. Many of my students are Chinese who received most of their education in China. What distinguishes them most from their American peers is an incredible preparation in mathematics.

How do you pack more math into a student? By raising standards and increasing difficulty of the curriculum. One thing I realized over the years is that success in mathematics is not just a natural ability. Instead, like in most other areas of life, success in math is 10\% ability and 90\% effort. Students who study more math and are regularly challenged with difficult content tend to reach a higher level. Students who are made to linger at the base to support their classmates’ grades lose motivation and then just lose.

As a parent, I know that the US math education standards are can be dramatically improved. And, yes, I have research on this, as well.

I would love to see President Trump and perhaps First Lady Melania make US Math education a high-priority item on the education scale. Doing so will also make the US labor force more competitive, reducing dependence on foreign workers. And, as the DeepSeek episode illustrates, in the age of AI, understanding math is more important than ever.

Irene Aldridge is a recognized quant researcher, entrepreneur, educator and author. Her most recent book is Big Data Science in Finance (2021, with Marco Avellaneda, Wiley, ISBN: 9781119602989).