Robotics & Geopolitics: India, China Forge AI Alliance; U.K. Weighs Robot Tax

Prime Minister of India Narendra Modi and President of China Xi Jinping in 2017. Credit: Kremlin.Ru

June 01, 2018      

As more countries invest in artificial intelligence, alliances between governments are starting to take shape. This week also saw another inquiry into the effects of automation on jobs in the U.K., which is considering a robot tax. In addition, one major business is taking a unique approach to detecting AI bias — by creating AI to detect it.

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?

China and India create big data, AI alliance to mend ties

Robotics development: A month after an informal summit between the leaders of India and China, the countries have launched joint projects in artificial intelligence and big data. The projects are being led by the National Association for Software and Services Companies (NASSCOM), India’s largest technology association, and the Chinese government.

On May 26, an AI project was launched in Dalian, China, and on May 27, a big-data project was launched in Guiyang, China. At the same time, to facilitate dialogue and projects between Indian and Chinese firms, a platform powered by AI called “Sino Indian Digital Collaborative Opportunities Plaza” (SIDCOP) was launched.

Geopolitical significance: India and China lead Asia for AI investment and adoption — for example, in India, AI investment and adoption jumped from 29% to 69% between 2016 and 2017. However, the AI markets of both countries have been unequal, independent, and disconnected.

A sign of this is the massive divide in investments: India’s entire AI budget is $480 million, while Alibaba alone is investing $15 billion to create AI labs around the world. In other words, a single Chinese firm is investing 30 times more than the entire Indian government. This imbalance makes an AI partnership between New Delhi and Beijing all the more noteworthy. Chinese money could pour into India, technology could be shared, and talent could move more easily.

Already, more than 50% of companies in India using AI are using it at scale, beyond trials and pilots. This number could grow even more as Indian firms tap Chinese AI.

But the big question is whether AI firms in the U.S., Japan, South Korea, and the U.K will lose out if firms in India gravitate towards Chinese AI. Consider that close to 70% of firms in India could deploy AI by 2022.

At the same time, it is uncertain as to whether this new AI partnership will be able to weather a future border dispute or other geopolitical disagreement. For now, it appears the next era of India-China relations is slowly being built upon AI.

Microsoft plans to use AI to remove bias in AI

Robotics development: Microsoft announced it is working on a “bias detection dashboard” that will find bias in AI that companies are deploying. The dashboard, which itself is AI, aims to stop people from being discriminated against.

If bias isn’t detected at an early stage, it could become harder to identify and could affect more people. At this point, it isn’t clear whether Microsoft will sell this tool to companies or keep it in-house.

Geopolitical significance: In addition to Microsoft, several other companies are taking steps to reduce AI bias. Facebook has a project called “Fairness Flow,” which can find bias in algorithms. It has already been tested in a job algorithm that Facebook deployed to connect job seekers with employers.

Bank of America is working on detecting bias as it deploys machine learning, such as for hiring people. The company said it doesn’t want to allow AI to have the final decision on hiring, and it claimed that AI bias can be fixed. At the same time, a consulting firm is offering businesses the chance to “test” how fair and bias their algorithms are.

Every AI firm must think about these issues. Even if its algorithms are being used in the most “mechanical ways” — such as crunching survey data and finding patterns — there could be bias. For example, is the AI allotting more weight to certain groups of people who took the survey?

What kind of bias then, could Microsoft’s tool find? One kind of bias is gender. For example, in an image-recognition program designed at the University of Virginia, areas like “kitchen” or tasks like “shopping” or “washing” were constantly linked with women.

Another example of bias could be against a country or foreign population? A U.S. firm may supply AI to help firms in Spain hire people. But the AI might constantly reject professionals from China because of the political climate in the U.S. Companies developing AI should think about country-level bias as they “export” their AI abroad. If they don’t, then perhaps IBM’s prediction will come true that by 2023, bias in algorithms may “explode.”

U.K. considers a robot tax to deal with automation job losses

Robotics development: The British House of Commons, through its Business, Energy and Industrial Strategy Committee, launched an inquiry into the effect that AI will have on work in the coming years. It will look at all aspects of AI transforming work and workplaces, including the automation of jobs, the effect on people, and ways to offset any imbalances, including a robot tax.

Geopolitical significance: Public policy will increasingly define how robotics and AI firms can operate and sell. For the U.K., this isn’t the first attempt to outline public policies around automation. In March 2016, a separate committee in the House of Commons launched an inquiry into AI and its implications. In July 2017, the British House of Lords launched an inquiry into AI and its ethical, moral, and social implications.

pixabay piggy bank robot tax AI alliance

Credit: AppletonOnfoot via Pixabay

Having multiple inquiries lets the British government collect ideas for how to deal with robotics and AI.

One of the ideas that may stick in the minds of policy makers is a robot tax. Why? Because millions of jobs are allegedly at risk of being automated in the U.K. The U.K. as a whole could see 30% to 35% of jobs disappear to robots. In Wales, one in three jobs could disappear by the 2030s. In Northern Ireland, 50% of jobs are at risk of being automated, and in Scotland 46% of jobs could be automated.

Even if these projections are just “fear mongering,” British policy makers may want to ensure they have a plan in place, and a robot tax could be a key part of such a plan. This could start a domino effect as other countries follow the U.K. and adopt a robot tax.

The U.K. itself may be late to the game. Last year, South Korea introduced what is being described as the “world’s first robot tax.” A robot tax might create havoc for AI and robotics firms. Companies that plan to automate certain processes may pause them to avoid paying higher taxes. This could cause robotics and AI firms to lose business. Do these companies have a plan to deal with this new disruption to their operations?

In addition, the governments that deploy robot taxes may face a different kind of problem. How do they keep firms that want to automate processes from leaving the country and moving to markets where there is no robot tax?