Chinese start-up DeepSeek’s remarkable AI software results have shaken public markets, driving down the shares of leading US technology companies, including Nvidia.
Many now wonder how a small team in China could, in such a short time, challenge the cutting-edge AI products of American companies with massive resources, outstanding talent and established market positions.
Small engineering teams working away from the mainstream have produced remarkable achievements in the past. Such events are unpredictable, but we should not be surprised when they occur, as I noted in a recent article.
This has been true in many fields. Alan Turing’s pioneering work in computer science innovation in Great Britain in the 1940s is but one outstanding example. China’s DeepSeek has thus joined a rare and distinguished group of small teams of pioneers who produce revolutionary advances.
Computer science is prime for such breakthroughs because high-power computing power is widely commercially available, and well-trained engineers are now active worldwide. Universities far and wide now provide outstanding engineering education. Meanwhile, open-source software has enabled wide access to sophisticated programming.
Equally important is the fact that leading technology companies from the US, including market leaders like Google, now develop significant parts of their software in locations around the globe using well-trained local engineers.
This creates domestic talent pools from which creative new companies can emerge to compete internationally. Once highly trained engineers are available, new software companies are likely to be formed, funded by governments and private capital.
Innovations will thus come from unexpected places— because brilliant people will do remarkable things given the right resources. And the resources for sophisticated software development are now available in many parts of the world.
The long-term effect of DeepSeek on the AI market is still being debated. However, it is safe to expect that the much lower cost of training useful AI systems will result in an expanded market as solutions become more affordable.
This has been the historical pattern in industrial technology, creating as usual winners and losers. But there is an important message in the success of DeepSeek for government policy on controlling technology transfer.
The widespread expansion of computer science skills will impact government attempts to control AI markets by restricting the nature of offered products. With creative international competition, there will be endless ways of bypassing restrictions as companies globally become effective competitors.
In the long term, no dependable barriers can protect products related to computer science. The equipment and technological skills know no long-term barriers – as Nvidia and Intel have demonstrated with opposite commercial value.
Nvidia took the initiative to enter the emerging AI market when it was risky. Intel did not. It happens that Nvidia is a US domestic company. Over time, new competitors will emerge from other geographies. Incumbents just have to be nimbler to prosper.
china produces more quality stem grad and phd than the entire world combined …
The US doesn’t make anything anymore and the things they make are crap. Look at their cars and planes. Total crapness.
China makes everything that isn’t crappy. Just look at their cars and planes.
How did we get?
There are two main things that will continue to advance technology breakthroughs: human brain power, and minerals that you extract from the ground such as rare earths. The second is limited in supply, but the first is unlimited. China excels in both despite US sabotage.