Artificial Intelligence - What Is The Asilomar Conference On Beneficial AI?


The Asilomar Meeting on Beneficial AI has most prominently portrayed social concerns around artificial intelligence and danger to people via Isaac Asimov's Three Laws of Robotics.

"A robot may not damage a human being or, by inactivity, enable a human being to come to harm; A robot must follow human instructions unless such orders would contradict with the First Law; A robot must safeguard its own existence unless such protection would clash with the First or Second Law" (Asimov 1950, 40).

In subsequent books, Asimov added a Fourth Law or Zeroth Law, which is often quoted as "A robot may not hurt mankind, or, by inactivity, enable humanity to come to harm," and is detailed in Robots and Empire by the robot character Daneel Olivaw (Asimov 1985, chapter 18).

Asimov's zeroth rule sparked debate on how to judge whether or not something is harmful to mankind.

This was the topic of the 2017 Asilomar Conference on Beneficial AI, which went beyond the Three Laws and the Zeroth Law to propose twenty-three principles to protect mankind in the future of AI.

The conference's sponsor, the Future of Life Institute, has posted the principles on its website and has received 3,814 signatures from AI experts and other multidisciplinary supporters.

There are three basic kinds of principles: research questions, ethics and values, and long-term concerns.

These research guidelines are intended to guarantee that the aims of artificial intelligence continue to be helpful to people.

They're meant to help investors decide where to put their money in AI research.

To achieve useful AI, Asilomar signatories con incline that research agendas should encourage and preserve openness and conversation between AI researchers, policymakers, and developers.

Researchers interested in the development of artificial intelligence systems should work together to prioritize safety.

Proposed concepts relating to ethics and values are aimed to prevent damage and promote direct human control over artificial intelligence systems.

Parties to the Asilomar principles believe that AI should reflect human values such as individual rights, freedoms, and diversity acceptance.

Artificial intelligences, in particular, should respect human liberty and privacy, and should only be used to empower and enrich humanity.

Human social and civic norms must be adhered to by AI.

The Asilomar signatories believe that AI creators should be held accountable for their work.

One aspect that stands out is the likelihood of an autonomous weapons arms race.

Because of the high stakes, the designers of the Asilomar principles incorporated principles that addressed longer-term challenges.

They advised prudence, meticulous planning, and human supervision.

Superintelligences must be produced for the wider welfare of mankind, and not merely to further the aims of one industry or government.

The Asilomar Conference's twenty-three principles have sparked ongoing discussions about the need for beneficial AI and specific safeguards for the future of AI and humanity.

~ Jai Krishna Ponnappan

You may also want to read more about Artificial Intelligence here.

See also: 

Accidents and Risk Assessment; Asimov, Isaac; Autonomous Weapons Systems, Ethics of; Campaign to Stop Killer Robots; Robot Ethics.

Further Reading

Asilomar AI Principles. 2017.

Asimov, Isaac. 1950. “Runaround.” In I, Robot, 30–47. New York: Doubleday.

Asimov, Isaac. 1985. Robots and Empire. New York: Doubleday.

Sarangi, Saswat, and Pankaj Sharma. 2019. Artificial Intelligence: Evolution, Ethics, and Public Policy. Abingdon, UK: Routledge.

What Is Artificial General Intelligence?

Artificial General Intelligence (AGI) is defined as the software representation of generalized human cognitive capacities that enables the ...