The Singularity Institute Blog

Earlier this year the OpenCog project was launched, with seed funding from SIAI and code and manpower donations from Novamente LLC. OpenCog is a free and open source software project aimed at providing a generic framework for the development of artificial general intelligence (AGI) software. The first formal code release is planned for Fall 2008, but the project has already gained considerable momentum. Most notably, through the Google Summer of Code project, Google Inc. has supported 11 student interns to work this summer on OpenCog-related projects. Several of these student projects have been extremely successful, yielding code and ideas that will form an important part of OpenCog going forwards.

Today another major step forward has been been taken, with an aim of pushing the OpenCog project in the direction of truly powerful AGI: the release of a wikibook outlining a design for a specific AGI system intended to be built on top of the OpenCog framework. This system design is called OpenCogPrime, and is heavily based on the Novamente Cognition Engine design under development at Novamente LLC during 2001-2008.

The OpenCogPrime design is proposed along with the hypothesis that, if the design is fully implemented and various important details are further refined, it may be able to form the basis of an AGI system with intelligence at the human level and beyond.

Of course, even in the case that this hypothesis is correct, it is difficult to estimate the amount of work that will required to create a human-level thinking machine according to the OpenCogPrime design. Levels of optimism among those involved with the project vary. My own (Ben Goertzel’s) personal intuition is that a human-toddler-level AGI could be created based on OpenCogPrime within as little as 3-5 years, and almost certainly within 7-10 years. The path from a human-toddler-level AI to an AI operating at the level of an adult human scientist is less clear, and could plausibly be even more rapid … or else much slower, depending on various factors (which are important and fun to consider, but would bloat this blog post too much…).

Clearly, there could be major unforeseen obstacles along the path to creating a powerful OpenCogPrime-based AGI; and it may turn out that OpenCogPrime is not a viable design for human-level AGI, for reasons that aren’t now anticipated by the system architects. But even if this is the case, we are confident that the process of refining, implementing, testing and teaching OpenCogPrime-based AGI systems will have a great deal to teach us about AGI and computing and cognitive science.

Those of us involved with OpenCog are extremely excited about the power of the free and open source development methodology for accelerating progress toward safe, beneficial AGI with intelligence at the human level and beyond. As in the case of Linux and other existing free and open source software projects, there is the opportunity for industry, academia and independent researchers and hobbyists to work together to create profoundly powerful software.

AGI ethics is (of course) at the core of SIAI’s mission, and we are dedicated to pursuing OpenCog and OpenCogPrime development in an ethically sound way. The free and open source methodology has obvious risks attached to it, and also obvious benefits: there will be a great number of intelligent, educated, concerned people looking at the code as it develops, with an eye toward ethical as well as intelligence-level considerations. Assuming OpenCogPrime-based systems successfully achieve greater and greater levels of general intelligence, we are committed to studying, monitoring and experimenting with these systems carefully with a view toward rapidly increasing our theoretical as well as practical understanding of the ethical properties of AGI systems of the OpenCogPrime category.

Onward toward an ethical, beneficial, progressively advancing and maturing artificial general intelligence!

Singularity Summit 2008 has been scheduled officially for Saturday October 25th (changed from the week before). It will be held at the Montgomery Theater in San Jose.

Watch this blog for further details, coming soon.

SIAI Research Fellow Eliezer Yudkowsky spoke on “(Artificial) Intelligence: The Wild Card” at the Global Catastrophic Risks conference, sponsored by the Future of Humanity Institute at Oxford University.

Yudkowsky’s essays “Cognitive biases potentially affecting judgment of global risks” and “Artificial Intelligence as a Positive and Negative Factor in Global Risk” will appear in the related book, Global Catastrophic Risks,” ed. Nick Bostrom and Milan Cirkovic (forthcoming, Oxford University Press, 2008).

Ronald Bailey covered the conference, including Yudkowsky’s talk, at Reason Magazine Online.

The Gazette, Montreal’s English-language newspaper, featured an article by David Sachs on the Singularity, “Survival of the Machines.”

The article mentions the Singularity Institute for Artificial Intelligence, and cites SIAI Director Ray Kurzweil, Director of Research Dr. Ben Goertzel, and Research Fellow Eliezer Yudkowsky.

Along with it appeared a shorter item “Friendly AI: More than taming robots,” specifically about Friendly AI.

Singularity Institute Research Fellow Eliezer Yudkowsky continues writing on the Overcoming Bias blog, sponsored by Oxford University’s Future of Humanity Institute.

Recent articles discuss free will, morality, and other topics related to SIAI’s research into safe artificial general intelligence.

Microsoft has acquired Powerset, a start-up with narrow AI technology for semantic search. Congratulations to CTO and co-founder Dr. Barney Pell, who is also a scientific adviser to SIAI.

SIAI Director of Research Dr. Ben Goertzel will speak at the World Future Society conference, to be held July 26 to 28 in Washington DC.

Dr. Shane Legg, a leading researcher of Artificial General Intelligence, has received the SIAI Canada Academic Prize for 2008. The $10,000 CDN prize is in recognition of his efforts to improve AI theory, and in the hope he will make further contributions in this field of study.

In his own words, this is what Shane is working on:

I’m currently enrolled at the University of Lugano in Switzerland as a PhD student in the department of Informatics. My thesis is written and submitted and I will be having my thesis defence in June. The title is “Machine Super Intelligence” in which I describe Marcus Hutter’s AIXI model and study some of its implications, extensions etc. [...] After my PhD is defended and I’ve finished my finance post doc, I am currently looking to go to London to do another post doc at an institute that specialises in theoretical neuroscience and machine learning. I plan to research hierarchical generative temporal models as I think this is one of the key properties of the neocortex, and thus general intelligence.

More information about Shane and his academic contributions can be found at http://www.vetta.org/

The SIAI Canada Association seeks to promote research in AGI, and related areas, and awards prizes to individual researchers, students and practitioners making original contributions to the field. A previous prize was awarded to Marcello Herreshoff in 2006. Suggestions of candidates for future prizes may be forwarded to canada@singinst.org
More about SIAI Canada can be found at http://www.singinst.org/aboutus/siaicanada

SIAI Director of Partnerships Jonas Lamis spoke with Doug Lenat, AI pioneer and founder and CEO of Cycorp. Lenat discussed how Cycorp can make progress in the next 10 years, why Moore’s Law is important, and why businesses should be interested in AI.

Video