Establishing an AI ethical standards will be more challenging than expected

Oct 28, 2018News

As people’s morals are different depending on their origin and class, some legal scholars argue ethics are too subjective for AI to follow in the conversation at the AI Now 2018 Symposium.

 

It is essential that AI need to follow standards to ensure the safety of human, the same goes to its developers, which has been emphasized by researchers, leaders, activists around the world. But who gets to define and enforce those ethics?

MIT Researchers have developed a platform known as the Moral Machine, which is used to look into the divergence of people’s opinions on driving situation and ask what they prioritize in an accident. The result from millions of people around the world shows that there is a huge variation across cultures. Moreover, it was proved by North Carolina State University that reading ethical standards actually doesn’t affect people’s behaviors, which means after reading a code of ethics, software engineers still does not do anything differently.

In the discussion, a solution was proposed to deal with the ambiguous and unaccountable nature of ethics: reframing AI-driven consequences in terms of human rights, suggested by an international legal scholar at NYU’s School of Law, Philip Alston. “[Human rights are] in the constitution,” said Alston. “They’re in the bill of rights; they’ve been interpreted by courts, if an AI system takes away people’s basic rights, then it should not be acceptable,” he added.

The Michael Dukakis Institute is currently collaborating with AI World to publish reports and programs on AI-Government, including AIWS Index and AIWS Products to solve key problems in ethical AI faced by businesses, governments and the world at large. Notably, this year the Michael Dukakis Global Initiative is the International Sponsor of AI World – aiworld.com – coming to Boston on Dec 3-5, 2018. AI World Conference & Expo draws over 3,000 businesses and technology professionals from Global 2000 enterprises, and includes 100+ sessions, 200+ speakers and 85+ sponsors and exhibitors. AI World is the largest independent enterprise AI business event in the world.  The conference program and special events are focused on how AI is helping to drive business innovation at large companies and covers the complete state of the practice of AI today. AI World is organized around this singular goal, enabling business leaders to build a competitive advantage, drive new business models and opportunities, reduce operational costs and accelerate innovation efforts.

Link of the AI World Conference & Expo Event: https://aiworld.com/