In September 2016 six leading AI companies — Amazon, Facebook, Google, DeepMind, Microsoft, and IBM — established a non-profit technology industry consortium called Partnership on AI (PAI). Their aim was to develop best practices in AI for social good in response to increasing concerns over the technology’s unprecedented impact on humanity. Apple joined soon afterward and the PAI consortium grew to include over global 70 partners. None, however, were from China.
Sinovation Ventures CEO and former Google China chief Kai-Fu Lee penned an open letter in protest: “China’s AI is also making rapid progress. It is unreasonable that [PAI] is exploring how to make rules and constraints for AI and prevent its negative impact on ethics and society without inviting Chinese companies to join in.”
Lee had a point, and this week PAI announced that Chinese search engine giant and AI innovator Baidu has joined the organization, marking the consortium’s entry into one of the world’s largest AI marketplaces.
Adding Baidu’s voice will broaden PAI’s understanding of global AI technologies and their ethical implications. Chinese companies are pushing forward with a wide adoption of AI technologies across industries from healthcare to transportation and beyond. This presents them with the opportunity to deal with practical issues that US companies and other global AI players may not have yet encountered.
Said PAI Executive Director Terah Lyons in the announcement: “Admitting our first Chinese member is an important step toward building a truly global partnership. The growth and scope of work on AI in China is extensive, and any conversation about the future of AI that does not involve China is an incomplete conversation.”
Perhaps more importantly, while both the US and China may have been reluctant to discuss AI ethics due their huge cultural and political differences, Baidu’s inclusion might break the ice and prompt more Chinese companies to get involved in such global discussions. For example, facial recognition technologies have become a universal identifier in China. But when Amazon reportedly offered its facial recognition tools to US police departments it encountered a significant backlash regarding civil liberties concerns.
Lyons told Synced via email: “We are not naive to the fundamental cultural and political differences between China and other parts of the world. Important cultural voices, activists, researchers, and commentators oppose elements of China’s political and societal structure, and we value those viewpoints.
“The pace of technological change in China is rapid, and the country is home to sophisticated AI technologists. There are important conversations about responsible development and use to be had; China’s context provides good reason to have them.”
It’s fitting that Baidu would be the first Chinese company to join PAI. Over the past few years it has been the most active and invested Chinese company in AI-driven fields. Baidu’s open-source AI platform Apollo, billed as “the Android of Auto”, now supports autonomous driving in geo-fenced areas. DuerOS, the company’s Alexa or Siri-style virtual assistant, provides over 90 million devices with voice recognition and conversational capabilities. Baidu Brain is the baseline for Baidu’s various AI services from image recognition to natural language understanding.
According to a person familiar with the matter, PAI and Baidu have been working on Baidu’s entry for some months. Baidu President Ya-Qin Zhang, an IEEE fellow and former lead of the Microsoft Research Asia team, has been a key figure in facilitating Sino-American collaboration on AI ethics.
“As AI technology keeps advancing and the application of AI expands, we recognize the importance of joining the global discussion around the future of AI. Ensuring AI’s safety, fairness and transparency should not be an afterthought but rather highly considered at the onset of every project or system we build,” said Zhang in the PAI announcement.
Baidu has been plagued by criticism that its search engine is over-commercialized or provides biased search results, and so the company has endeavored to preemptively limit negative consequences before wide implementation of its AI technologies to avoid a backlash in that arena. At the World Artificial Intelligence Conference in Shanghai this month, the search engine giant’s Founder and CEO Robin Li promoted his four principles of AI ethics: Be safe and controllable; Promote more equal human access to technology; Help humans learn and grow rather than replacing them; Bring more freedom and possibility to mankind.
PAI has three Working Groups: AI, Labor, and the Economy; Safety-Critical AI; and Fair, Transparent and Accountable AI. Baidu says it will “contribute to research projects that are intended to develop the policies, tools, and principles that will inform and drive responsible AI development and deployment.”
Journalist: Tony Peng | Editor: Michael Sarazen