AI Nationalism & The Fight To Win The AI Wars
AI Nationalism is a concept that I have been thinking a fair amount about these days as principal investors in the machine learning and artificial intelligence (AI) ecosystems. With many companies in our portfolio focused on machine learning and artificial intelligence, I spend time thinking through the impact of these companies as it relates to labor markets, income generation, the future of work, and automation. Something that comes up when you go down this path is the impact of AI on nation-states and what the continued progress in machine learning means for geopolitics. Many people have already coined this idea via a term: AI Nationalism. My personal view is that AI policy will likely become one of the most important areas of government policy as this rapid progress occurs and an “AI arms race” likely takes shape. While this concept seems so crazy in light of the recent Mark Zuckerberg hearing it seems inevitable as advancement continues. According to Element AI, there are fewer than 10,000 people that have the skills necessary to tackle serious artificial intelligence research. Contrast that with over seven billion people on earth and you can see how unique this is. In short, a single general purpose technology will likely have a huge impact on so many parts of national policy. Continued advancement in machine intelligence, process automation, and artificial intelligence will likely drive a massive increased focus on this area of policy and work. Some countries will stay far ahead of the pack, others will be in the pack, and some will be very far behind.
What Do We Mean When We Say This?
Nationalism is a political, social, and economic system characterized by promoting the interests of a particular nation particularly with the aim of gaining and maintaining self-governance, or full sovereignty, over the group’s homeland. Artificial Intelligence is intelligence demonstrated by machines, in contrast to the natural intelligence (NI), displayed by humans and other animals. In computer science, AI research is defined as the study of any device that perceives its environment and takes actions that maximize its chance of successfully achieving its goals. Colloquially, the term “artificial intelligence” is applied when a machine mimics “cognitive” functions that humans associate with other human minds, such as “learning” and “problem solving.” Combined, what I mean when I say this is a nationalistic fervor around the subject of AI as governments seek to rally support for their vision. The combination of AI evolution and nationalistic tendencies will lead to the conclusion that the winner in the AI race will wield extraordinary power.
Where Are We At Today?
To set some framework, the US is widely regarded as the leader in AI today both in terms of research and industry. However, beyond a 2016 Obama administration report called Preparing For A Future In Artificial Intelligence we do not have an overall industrial policy or plan. The private sector mostly drives the success of the US. The Chinese government released a report called A Next Generation Artificial Intelligence Development Plan in 2017 that laid out an overall industrial policy and plan. In short, their goal is to dominate the AI industry by 2030. They followed this up with a plan to spend $2.1B on an AI industrial park near Beijing that has room for 400 companies. The French government released a report called AI for Humanity in 2017 that laid out an economic strategy for developing France’s AI ecosystem. In addition, the government pledged $1.8B into AI by 2022 to grow their ecosystem. Right now, it is basically a fight to shape the future with China and France having explicit investments and national strategies and the US leading mostly due to enormous private company investment in AI. There are other nations I do not include here however these are the three that are farthest along.
China, China, China.
When you think about AI Nationalism, you probably cannot invent a more perfect example of how to develop a market leader than China. Protectionism did not really allow any outside actors to encroach on domestic industry. Communism allowed the state to set policy and framework and invest where they wanted without much debate. Domestic private and public companies become very coupled to the national strategy. Lax privacy laws allow domestic companies to amass enormous amounts of data. In short, China will be the one to watch as it relates to AI Nationalism. As I mention above they have an explicit goal of becoming the AI leader. They are investing billions in research and companies. The population size, lax privacy laws, and huge adoption of new technologies allows them to amass orders of magnitude more data than others (data and computing power are the most important asset in AI). New cyber laws require a review of data exports across Chinese borders. The Chinese government is guiding national champions in key AI fields. Andrew Moore (Dean Of CS at Carnegie Mellon) testified to the US Senate in 2016 that the percentage of papers submitted from China to AI conferences increased from 5% a decade ago to 50% today. Note that this is what they are sharing with the world and likely not all of the papers.
Public vs. Private Sector.
Most of what I have been talking about is a focus on public sector work. However, one cannot think through AI Nationalism without thinking about the private sector. Companies like Google, Tencent, Amazon, Facebook, Apple, Alibaba, and Baidu are making huge investments in AI and machine learning. It is safe to say their expertise in this arena is greater than any state actor at the moment. These private companies also have much deeper pockets (and access to better computation), so they can often times run incredibly expensive experiments to produce high-impact outputs. Something to think about is whether or not any of these companies can maintain independence. Do you believe that Google will be able to stay independent of alignment with the US if we have an AI arms race (for instance)? Could the employees have grassroots impact to force a private company to “not participate” in these areas? These are interesting questions to ponder and if it is anything like the prior arms race the blurring of lines between the private and public sector is going to be important.
Digital Geneva Convention.
There has been a lot of discussion around this idea with some private companies coming out in support of a Digital Geneva Convention. The basic idea is that we now need a Digital Geneva Convention that will commit governments to protecting civilians from nation-state attacks in times of peace. This remains a widely discussed idea but the question largely remains whether or not you can actually get everyone to talk to each other, share information, and debate. From there, can you get everyone to develop norms, protocols, and verification mechanisms? This seems too optimistic to me and I don’t believe that this will ultimately happen over even a long period of debate and discussion. The most interesting actors I think are organizations like OpenAI and what they are doing to help push policy and capability in this area.
Nation State Allegiances.
Allegiances get interesting at the nation-state level. I am not talking about the relationship between public and private companies, but more the relationship between nations and their allies. I believe there is going to be a component of this where someone like the US or China (for example) becomes so powerful in AI that they share their playbook, approach, and output with other countries. They share their talent. They share their services. Sort of “we will give you our playbook in exchange for for allegiance and interconnectedness.” I wonder if you see this sort of allied and axis powers similar to World War II based on a set of principles (military, technology, economy), help, subordination, and understanding.
AI Creating Isolationistic Bias.
In the era of this impending “AI arms race,” a new problem will likely emerge: isolationistic AI bias and AI ethical challenges. Ultimately, it is critical to remember that AI is driven by what models are trained on, what models are trained for, and how models are applied. Humans are deeply involved in not only framing, but the application of AI. So, the bias of AI reflects human bias, largely implicit and unintentional. Diversification and global collaboration are critical to combating the gender, racial, and income biases that are inherent to AI in its current form and corporate development. AI Nationalism, with isolated research, can intensify this bias. Bias could become so substantial that frameworks and systems developed in one country cannot be applied to another country because of differences in culture and racial demographic.
Those are some of my thoughts on the concept of AI Nationalism and where this ultimately could go. The impacts economically, militarily, and technologically are profound and it is something that I have been thinking a lot about. I plan on continuing to update this blog as more information and thoughts flow in.
If you liked “AI Nationalism & The Fight To Win The AI Wars” and want to read more content from the Bowery Capital Team, check out other relevant posts from the Bowery Capital Blog.