Artificial Intelligence

The mission of the Foresight Institute is to promote the beneficial development, and help avoid the dangers, of nanotechnology, AI, biotech, and similar life-changing technologies. Most of our attention has been focused on nanotechnology, especially as derived from Richard Feynman 's 1959 vision of the construction of atomically-precise products through the use of molecular machine systems. Artificial Intelligence (AI), however, like nanotechnology, provides important approaches for addressing many of today's most important problems, while the most advanced and far-reaching development of the technology—atomically precise manufacturing in the case of nanotechnology; artificial general intelligence in the case of AI—will present both vast new opportunities and significant risks.

Cyber, Nano, and AGI Risks: Decentralized Approaches to Reducing Risks

New - March, 2017

"Cyber, Nano, and AGI Risks: Decentralized Approaches to Reducing Risks" Published in Proceedings - First Colloquium on Catastrophic and Existential Risk Garrick, B.J. (Editor), sponsored by The B. John Garrick Institute for the Risk Sciences, Henry Samueli School of Engineering and Applied Science, University of California, Los Angeles, UCLA Luskin Convention Center, March 27-29, 2017.

See also "Cyber, Nano, and AGI Risks: Computer Security and Effective Altruism".

Thinking Machines in Foresight's Founding Vision

Although the book that presented Foresight's Founding Vision likewise focused on nanotechnology, artificial intelligence was discussed as another transformative technology in the early stages of development, presenting great opportunities, but also dangers to be avoided. "Thinking Machines" was the title of Chapter 5 of K. Eric Drexler's 1986 book Engines of Creation.

To show how machine intelligence could develop, Drexler outlined how human intelligence evolved by variation and selection from molecules that behave to promote their replication, to bacteria that move guided by trial and error, to worms that evolve behavior produced by evolving genes, to animal brains that learn to think by evolving internal models of the external world. Human language enabled mental models to be shared, to spread, and to evolve standards for judging the worth of different mental models. Minds thus evolved to form societies of agents that compete to establish their fitness for specific purposes. Mental agents formed by teaching and imitation are called "memes". Minds thus expand knowledge by varying idea patterns and selecting among them. The natural process by which intelligence evolved points to how intelligent machines might work.

After outlining the progression of machines that think from Charles Babbage's mechanical calculator of the mid 1800s to the expert systems of the 1980s, Drexler concludes that, just as ancient speechless human ancestors brought forth by genetic and cultural evolution literate human societies, evolving memes of technology will bring forth machines that think—machines able to learn and organize knowledge.

Drexler distinguishes two types of artificial intelligence: technical AI (adapted to deal with the physical world), and social AI (adapted to deal with human minds and pass the Turing test). Drexler focuses on technical AI—an automated engineering system—as the more important for its impact on the technology race, in particular, its impact upon the race to develop advanced nanotechnology (atomically precise manufacturing). He then describes the success of an early automated design system named EURISKO.

With this background established, Drexler casts the importance of advanced AI in terms similar to his treatment of the opportunities and dangers to be brought by advanced nanotechnology:

In a sense, artificial intelligence will be the ultimate tool because it will help us build all possible tools. Advanced AI systems could maneuver people out of existence, or they could help us build a new and better world. Aggressors could use them for conquest, or foresighted defenders could use them to stabilize peace. They could even help us control AI itself. The hand that rocks the AI cradle may well rule the world.
—K. Eric Drexler Engines of Creation

And also:

… Technical AI is emerging today [1986], and its every advance will speed the technology race. Artificial intelligence is but one of many powerful technologies we must learn to manage, each adding to a complex mixture of threats and opportunities.
—K. Eric Drexler Engines of Creation

Relationship of design ability and fabrication ability

A Foresight Background essay "Dimensions of Progress" published in 1987 explores the effect of developing atomically precise manufacturing on the development of AI, and the effect of developing AI on developing APM.

… The transition to nanotechnology will look different if it is aided by Al designers; the transition to Al will look different if it occurs on a foundation of abundant assembler [nanofactory]-built hardware.
—K. Eric Drexler "Dimensions of Progress"

The Synergy of Molecular Manufacturing and AGI

Former Foresight President (February 2009-March 2010) J. Storrs Hall wrote popular books about both nanotechnology (Nanofuture: What's Next for Nanotechnology (Prometheus 2005), which won the Foresight Communications Prize and the Bela Kornitzer prize) and AI (Beyond AI: Creating the Conscience of the Machine (Prometheus, 2007), the first full-length non-fiction treatment of machine ethics).

The 2010 Foresight technical conference (January, 2010) celebrated the 20th anniversary of the First Foresight Conference on Nanotechnology (October, 1989). Foresight 2010: the Synergy of Molecular Manufacturing and AGI also focused on the complex synergy between advancing nanotechnology and advancing AI, and the possibility that progress during the coming few decades in manufacturing (advanced nanotechnology) and intellectual work (advanced AI) might equal the progress seen over the previous few decades in computing and networking capability. J. Storrs Hall's perspective on how nanotechnology had changed in the 20 years since the first Conference also illuminates the potential illustrated at the 2010 Conference for exponential technological advances leading to both nanofactories and self-improving AIs.

Ethical Issues in Artificial Intelligence

Top 9 ethical issues in artificial intelligence

October 21, 2016—Foresight President Julia Bossmann has an article on emerging technologies on the World Economic Forum web site "Top 9 ethical issues in artificial intelligence" ranging from issue #1 - "Unemployment" to issue #9 "Robot rights".

On Foresight's blog: Nanodot

Since its beginnings in May of 2000 Foresight's blog Nanodot has occasionally dealt with "Machine Intelligence" and "Artificial Intelligence":