Self-Aware Systems is a think tank that develops innovative approaches to a more productive and cooperative human future. We are at an amazing and critical juncture in human history. We face many challenges: economic instability, overpopulation, shortages of water, oil, and raw materials, disease, terrorism, war, and the destruction of species and ecosystems (eg. see Jeffrey Sachs, “Common Wealth: Economics for a Crowded Planet“). But we are also in the midst of scientific and technological revolutions that may dramatically improve our lives. It is remarkable that simultaneous breakthroughs are occurring in biology, neuroscience, psychology, artificial intelligence, nanotechnology, and fundamental physics. If we navigate this period of remarkable change with wisdom, we have the potential to create a compassionate and empowering future.
Self-Aware Systems aims to promote cooperation at all levels of interaction through careful design of new technologies. The same fundamental cooperative principles apply to biological systems, psychological dynamics, group interaction, business dynamics, economic systems, ecosystems, and political institutions. As interactions become increasingly computational, we have a tremendous opportunity to create more beneficial outcomes by designing new interaction rules and institutions. For example, new incentive structures can promote more productive cooperative behavior in business and new computation-based social structures can promote more peaceful and productive global interactions. Some of the technologies involved include incentive design, formalized contracts, formal provenances for trust establishment, provably limited monitors, oblivious computation, and commitment mechanisms based on formal goal and source code revelation. Many of these technologies have both immediate applications and serve to create the infrastructure needed to build an advanced cooperative society in the future.
Talks and White Papers
As a part of our mission, we are producing talks and white papers on certain aspects of our work to stimulate a broader discussion of important issues.
This two-hour long talk discusses the science and technology of cooperation:
Creating a Cooperative Future
This hour-long talk discusses the future of technology:
Stanford Computer Systems Colloquium Talk: Self-Improving AI and the Future of Computing
This 10 minute talk summarizes the AI drives argument:
AGI-08 Talk: The Basic AI Drives
This half hour interview sets the context for the work:
SIAI Interview with Steve Omohundro
This hour and a half talk discusses connections with morality:
AI and the Future of Human Morality
This two hour talk discusses connections with biology:
Co-opetition in Economics, Biology, and AI
This hour long talk discusses choices for the immediate future:
Foresight Vision Talk: Self-Improving AI and Designing 2030
For more details, read the two papers:
The Basic AI Drives
which one reviewer described as “the best piece of work on AI goals and behavior I saw in 2007″ and the more technical analysis:
The Nature of Self-Improving Artificial Intelligence
Send email to Steve Omohundro:
