One might think, in order to become a leading nation in AI technology, you need the best AI engineers and scientists. It certainly is a factor, but not the single predominant one. But which other factors should we be looking at?
As soon as you have worked on both a project, where you trained machine learning models with a small dataset, and a project with a large dataset, you know how important the data is for your algorithms and models to perform well. With almost abundant data, you will most likely be able to do quite well with simpler models and algorithms, whereas if you have very little data you need to optimize pre-processing and dig deep down into your box of tricks. It’s probably safe to say, that a good AI engineer with abundant data can almost always outperform a top-notch AI engineer with little data, assuming that the quality of data is comparable. Therefore, being able to gather a lot of high-quality data is certainly another major factor to consider.
This brings us to another very important factor, a supportive policy environment, which stands for example for subsidizing the technology and also sets rules for gathering and using data. As mentioned in Part I, regulations are different for every country and therefore the ability to gather data is different. An AI engineer with an idea, but imposed by restrictions to use the needed data, will most likely just try to realize that idea somewhere he is allowed to do so. Launching a startup or joining an accelerator program abroad has become fairly easy and therefore is definitely an option for a person determined to make his/her idea happen. Another dimension of policy environment is for example subsidizing AI development on different levels, as described by Kai Fu Lee in his book AI Superpowers, where the Chinese government is ”offering subsidies for research, directed venture-capital guiding funds, purchasing the products and services of local AI startups, and set up dozens of special development zones and incubators” as well as “simplified procedures for registering a company, free shuttles, special apartments” and many more. This really shows the amount of effort China is putting into AI development and how important it is.
While research is very important to bring new technologies forward, it is equally important (or you might argue even more important) to translate this into industry and daily lives. China is an extremely lucrative and attractive market, as launching a product will potentially reach a market of over a billion people. Further, the Chinese market is very competitive and has less regulations than western countries, which leads to a very interesting battle amongst the companies aiming for the same markets and products. Kai Fu Lee describes the Chinese entrepreneur and their mentality as a gladiator and that “the only recourse when an opponent strikes a low blow is to launch a more damaging counterattack”, which produced many energetic and “street-smart” entrepreneurs, who have become very successful at bringing AI technologies to the market. An analogy to chess seems appropriate as well, strategically figuring out which moves to take, which sacrifices to make in order to still achieve the overarching objective of defeating the opponent and emerge as the successor. The AI chess game has started.
Concluding, some of the main factors to become a leading nation in AI are a supporting policy environment, smart and strategic entrepreneurs bringing AI to industry and embedding it in daily life, a large amount of good quality data and of course well-trained AI engineers. As seen, a lot of these factors can be influenced to a substantial amount by governments.