Robotics & Geopolitics: AI Strategy Must Hinge on Local Strengths, Markets

Credit: The Digital Artist, via Pixabay

May 25, 2018      

A nation or region’s artificial intelligence strategy is viewed as essential to future economic and geopolitical competitiveness, but we can see a number of approaches. Making sure that machine learning is localized, centralized hubs, and targeting niche markets, are among them. How are China, the U.S., and other nations developing — or not pursuing — an AI strategy?

Robotics Business Review has partnered with Abishur Prakash at Center for Innovating the Future to provide its readers with cutting-edge insights into recent developments in international robotics, AI, and unmanned systems. Are you ready to be updated?

Regular people, not programmers teach robots

Robotics development: Nvidia’s robotics research lab has developed neural networks that allow robots to learn new skills by watching and monitoring human behavior in an environment like a factory.

The power of this innovation is that it could allow robots to move beyond repeating actions. They could start to adapt to changing conditions, such as if a robot or a human drops a box on the factory floor. In addition, robots could be personalized by changing their behavior based on what they observe their owners doing.

Geopolitical significance: One of the big challenges for organizations that want to use advanced automation is the time it takes to train AI. The way a bank wants AI to behave is very different from how a coffee shop wants a robot to behave.

As Nvidia solves this challenge, it could drive automation around the world. But how will technology suppliers ensure that their AI strategy adapts to different cultures?

Indian Prime Minister Narendra Modi is a proponent of having an AI strategy.

Prime Minister Narendra Modi wants Indian agriculture to use AI. Credit: Sambeet, via Pixabay

For example, workplace automation in India is expected to double in the next three years. One sector in India that may turn to automation is agriculture. Narendra Modi, India’s prime minister, is actively promoting AI in agriculture, and a government-appointed committee identified agriculture as one of 10 areas where India should invest in AI.

However, agriculture in India is different from agriculture in the U.S. Will human-directed machine learning take this into consideration when making decisions?

There is the risk that AI that learns from humans could fail. Last year, Nvidia signed a deal with Toyota to use Nvidia chips in Toyota’s self-driving cars. In January, 2018, Cherry, a major automaker in China, signed a similar deal with Nvidia.

However, the driving culture differs widely between Japan and China. If AI is learning from human drivers in Japan and China and updating itself, it could make a dangerous mistake if it applies lessons learned in one place while operating in the other.

Companies developing AI that learns from people must include cultural differences, otherwise they could be blamed when their AI fails abroad.

China designates Hong Kong as an AI hub

Robotics development: Alibaba, SenseTime, and the Hong Kong Science and Technology Parks Corp. (HKSTP) are establishing the Hong Kong AI (HKAI) Lab. The organizations hope to turn Hong Kong into a regional hub for AI by attracting startups. Starting in September, HKAI Lab will offer a six-month accelerator program for AI startups.

Geopolitical significance: Unlike Western technology firms, which are not directly controlled by their respective governments, when Chinese technology firms take a step in a particular direction, there is always an elephant in the room: How much has Beijing influenced the decision?

Turning Hong Kong into a regional (or global) hub for AI could heighten these concerns for countries like the U.S. and Germany, which want to protect their technology from China. In February, the U.S. blocked a Massachusetts-based electronics company from being acquired by a fund controlled by Beijing.

June 18-19, 2018 -- Boston, MA
Join other executives and professionals striving to harness innovative technology into modern manufacturing organizations.

And in April, it was reported that the U.S. government could start “scrutinizing” any “informal partnerships” between American and Chinese AI companies. As the U.S. bans companies like ZTE and Huawei, could the U.S. AI strategy prohibit companies from receiving Chinese funding — including through programs such as the HKAI Lab? This was proposed last year.

If this happens, it could hurt Hong Kong’s aspirations to become a hub for AI. However, it could also push China to stop its firms from getting support from the U.S. American firms like Google, which recently opened an AI lab in China to win support from Beijing, would be affected by this.

AI and robotics firms in the U.S. should begin thinking about where they will receive funding from if the U.S. blocks Chinese investments. Would this derail the American AI industry? Will AI firms from other countries have a leg up over the U.S. as they gain access to Chinese funding?

What kind of AI strategy should the U.S. have?

Robotics development: John K. Delaney, an American congressman and a U.S. presidential candidate for 2020, has questioned why the U.S. doesn’t have an AI strategy even though France, China, and the European Union (EU) do.

To move the U.S. closer to a formal AI strategy, Delaney and other US politicians have introduced a bill called the “Future of AI Act.” Through this bill, a committee would look at different ways machine learning and related technologies could impact society, business, and more. The committee will then create recommendations for the U.S. government.

Geopolitical significance: The U.S. may want an official AI strategy, but France and the EU may be the wrong models. France is investing just $1.8 billion over five years to become an “AI power.” Consider that Sinovation Ventures, a fund in China, has unveiled a $900 million AI budget.

In other words, a single, private fund has 50% of what the entire French government wants to invest over five years. At the same time, the EU has no formal AI strategy. While 25 EU members have signed a declaration around AI and the European Commission wants 20 billion euros to invest in AI, none of these strategies are resulting in much.

Instead of France or the EU, the U.S. may want to look at the U.K. The U.K. has been actively investing in AI (and other technologies). Last November, it unveiled $633 million in funding.

In April 2018, the British government announced an additional $1.8 billion. However, the U.K. knows it can’t match China or the U.S. in funding, so it is trying to find its niche. Two niches are emerging: ethics and healthcare.

Could the U.S. identify niche areas within AI to fund? One niche area could be AI cybersecurity. After all, the U.S. signs the most cybersecurity deals in the world — 69% of all deals. As contradictory as it may sound, such a niche could be how the U.S. competes holistically and dominates this new geopolitical competition.