Will AI be guided by Western or Asian values?

Will AI be guided by Western or Asian values?

As China and the United States race to become the preeminent artificial intelligence (AI) superpower, an interesting question arises: Will the AI of the future be dominated by Western or Asian values?

All algorithms are encoded with a set of human values as the foundation of AI, and this has up until now been driven by Western values such as individual rights, privacy and transparency. However, this may change as China rapidly develops its own AI capabilities including its own set of ethical standards.

In May, the Beijing Academy of Artificial Intelligence released the Beijing AI Principles developed by China's most prominent technology organisations. It stressed the importance of using AI for the benefit all of humankind, along with issues of accountability and security, and the need for harmony and cooperation in the development of an AI ecosystem.

The publication of China's principles came just one month after the release of the EU's Ethics Guidelines for Trustworthy Artificial Intelligence. If we compare the two sets of principles, what do we see? Common features such as technical robustness and accountability, diversity and inclusiveness, human oversight and benefits for humanity.

As for the differences, China's principles include having an open AI platform, education and interdisciplinary cooperation, and considering the impact of AI on human employment. Points stressed more strongly in the EU standards are the need to be "sustainable and environmentally friendly" and a "full respect for privacy". China's values aim for the collective good, whereas the latter possesses a more individualistic worldview.

In the US, the technology giants driving the development of AI have developed their own ethical principles. The released last year by Google include: to be socially beneficial, avoid creating or reinforcing unfair bias, and accountability to people.

Google's principles also strongly emphasise privacy, avoidance of surveillance that "violates internationally accepted norms", and unfair bias in areas such as ethnicity, sexual orientation or religious beliefs.

While all sets of principles mentioned include the importance of being socially beneficial, this of course is subjective and depends on the prevailing societal standards, which may not be the same in different cultures. For instance, attitudes toward privacy and security, individualism and community tend to differ between Asian and Western societies.

In the future almost every area of our lives will be affected by decisions made by the algorithms that drive AI. It is therefore encouraging that there are more areas of agreement than difference in the efforts to develop common values and principles. As Thailand is adopting major technologies from both the West and the East, we should pay close attention to how this develops.

Suwatchai Songwanich is the CEO of Bangkok Bank (China). For more columns in this series please visit www.bangkokbank.com

Do you like the content of this article?