Let’s Talk AI Cooperation (Don’t Call it an “Arms Race”)

Advances in artificial intelligence are usually framed as a winner-take-all competition. Instead, let’s talk about AI cooperation. In this article, we’ll examine AI breakthroughs through a sunnier lens, and discuss the danger of framing AI progress as an “arms race”.

The AI “arms race” competitors will actually need each other to keep developing AI technologies

The supposed “artificial intelligence arms race” between the United States and China is a popular subject. Wikipedia even has an entire article on it. The story goes that the countries fear that new AI technologies will grant some insurmountable advantage in a potential war. And so, they’re doing everything they can to beat the other.

For example, the Chinese government has plans to establish a state-owned $30 billion venture capital fund for the advancement of artificial intelligence systems. Compare this to the U.S. Defense Department’s budget of merely $2 billion for their “AI Next” initiative through DARPA. On the other hand, the U.S. has a stronger private sector presence in AI. In fact, as of 2018, the U.S. claims 40% of the “AI market share” (1,393 companies).

Interestingly, experts are starting to note that the future will be very different. The U.S. and China, they say, will actually depend on each other for ongoing AI development.

To quote Pascale Fung, director of the Centre for Artificial Intelligence Research at the Hong Kong University of Science and Technology,

“To prevent an AI dystopia and to maximize the benefits of AI, China and the U.S. need to continue working together, even if not at the government level, at the academic level, at the company level in the private sector.” (Source)

In fact, Fung predicts that, “You will see a lot more formal and open collaboration between Chinese and American researchers.”

Fung emphasizes a perspective that we rarely see: that of cooperation and collaboration. Her words stand in stark contrast to the Cold War-era motifs of “arms races” and “competitions”.

AI cooperation between the U.S. and China isn’t just inevitable, it’s already happening

The reality is that China and the U.S. have an increasingly complex bilateral relationship. Between them, artificial intelligence is emerging as a new ground for AI cooperation. At the national level, the U.S. and Chinese governments both lean towards arms races and competitions. But the private business and technology sectors of their economies are becoming more and more entangled.

Microsoft, for example, recently launched an AI-powered public services platform in Qingdao, China. The tech giant worked with Huashi Cloud Education and the Qingdao government to build a virtual reality industrial center and AI industrial park.

Microsoft executives and Qingdao government officials meet to launch their new public services platform. Image source

According to Alain Crozier, Corporate Vice President, “What we are doing together here in Qingdao will help the 93% (of companies) thrive.”

Don’t call it an “AI arms race”

Using the term “arms race” to describe AI development makes artificial intelligence into a one-dimensional beast. It implies that countries are developing the technology on their own without any cooperation – which, as we’ve established, isn’t true.

Worse, “arms race” immediately makes you think of the Cold War and using AI to kill people. And yes, the Pentagon is “doubling down” on using AI for national security (which, of course, means killing people better).

In fact, Google, which has a big AI lab in Beijing, recently received some heat from the Pentagon over their AI presence in China. General Joseph Dunford, chairman of the Joint Chief of Staff, aired his concerns that the Chinese military could benefit (directly or indirectly) from Google’s artificial intelligence work in their country.

However, the overwhelming majority of AI is used for non-military applications. As Microsoft VP Alan Crozier also pointed out at the launch of their public services platform in Qingdao, “Vital industries like agriculture, manufacturing, natural resources and shipping are transforming and need to use [AI] technology to become smarter.”

International AI cooperation also comes through universities and educational partnerships

Tech giants like Microsoft and Google aren’t the only ones fostering international cooperation in AI research. American and Chinese universities, including Stanford and Tsinghua, are also on the forefront of global cooperation in AI R&D.

Stanford University, widely-known and located in the heart of Silicon Valley, has an ongoing research partnership with Chinese corporate giants DiDi and Huawei. Tsinghua University is noted for their corporate partnership with Alibaba, the multi-national tech conglomerate.

Last year, Tsinghua announced that they are opening the new Tsinghua University Institute for Artificial Intelligence. Through the Institute, the university will work with Google and Tencent to develop new AI algorithms and hardware. Tsinghua University Advisory Committee on Computer Science even brought on Google’s AI Chief Jeff Dean to lead the charge.

Cities are also partnering with AI companies

And the story doesn’t end with universities and private tech companies. Individual cities are creating deliberate AI business ecosystems to draw international investment and research talent. Montréal, for example, now hosts AI research labs for Microsoft, Facebook, Samsung and Google, and ElementAI, as well as myriad smaller companies using AI to build better technology solutions.

Yoshua Bengio, co-founder of ElementAI and joint recipient of the 2018 Turing Award, says he believes that, “Montreal has emerged as a powerhouse due to the combination of great universities, great companies and Canada’s ethos of cooperation among elite minds.”

Some cities are partnering directly with AI companies to build tech solutions to infrastructure problems and improve quality of life. ElementAI is about to launch a solution for the Port of Montréal to improve the movement of trucks in and out of the port. And SingularityHub outlines 8 more ways that AI will transform cities as we know them.

The future of AI is global cooperation

Moving forward, will we see more AI collaboration or a return to arms races? Chinese President Xi Jingping is at least saying some positive words. In a letter to September’s 2018 World Artificial Intelligence Conference, President Jingping wrote:

“We’re hoping that all countries, as members of the global village, will be inclusive and support each other so that we can respond to the double-edged-sword effect of new technologies. AI represents a new era. Cross-national and cross-discipline cooperation is inevitable.” (Source)

Honestly? As demand and development of AI continues to grow, we expect to see both more collaboration and more competition. The truth is that we have yet to work out a perfect model for international collaboration in any field, let alone AI. The algorithms and AI tech itself are actually shared quite freely by tech companies and researchers. But the data used to train the AI tends to be jealously guarded.

Ultimately, it will be up to the luminaries and leaders of private institutions to lead the charge for greater AI cooperation in the decades to come.

Related articles and further reading

3 Surprising AI Applications in Food, Energy and Airlines

Artificial Intelligence for Disaster Relief

Pentagon outlines its first artificial intelligence strategy

Deepfakes Explained: What, Why and How to Spot Them

Artificial Intelligence at the FBI – 6 Current Initiatives and Projects