OpenCyc project: Open sourcing knowledge for AI | Douglas Lenat and Lex Fridman

TL;DR

Open Psych aims to build a knowledge base that is open to the public while respecting the proprietary information of its clients. The company struggles with deciding what parts of their system to open source and collaborate with the community on.

Transcript

one of the ways that people can help is by adding to the knowledge base and that's really basically anybody if the tooling is right and um the other way i kind of want to ask you about your thoughts on this so you've had like you said in government and you have big clients you had a lot of clients but most of it is shrouded in secrecy because of th... Read More

Key Insights

• ❓ Open Psych separates the problem of determining what the system should know from the problem of how the system can efficiently reason with that knowledge.
• ❓ The company has struggled to convince AI researchers of the importance of their expressive representation language and finds that many researchers settle for a simplified version.
• 🎚️ Open Psych has developed heuristic level modules to improve the efficiency of reasoning, with over a thousand modules contributing to finding solutions.

Questions & Answers

Q: What is the main rationale behind Open Psych's approach to handling proprietary and public knowledge in their knowledge base?

Open Psych respects the proprietary information of its clients and marks it as such in their knowledge base. However, publicly available, general human knowledge becomes part of the open Psych knowledge base.

Q: Has Open Psych faced challenges in convincing AI researchers to adopt their expressive representation language?

Yes, Open Psych has struggled to convince AI researchers of the importance of their expressive representation language. Many researchers have made do with a simplified version of their system instead.

Q: How does Open Psych utilize heuristic level modules to improve the efficiency of the system's reasoning?

Open Psych has over a thousand heuristic level modules that function as a community of agents. These modules contribute to reasoning tasks and gradually find solutions to problems by leveraging the knowledge base.

Q: How does Open Psych approach the decision to open source parts of their system and collaborate with the community?

Open Psych carefully considers what to open source and what to keep proprietary. They often make the entire knowledge base and executable code available to collaborators but provide limited access to source code.

Summary & Key Takeaways

• Open Psych allows public access to general human knowledge in their knowledge base, while respecting the proprietary information of their clients.

• The company separates the epistemological problem of what the system should know from the heuristic problem of how the system can reason efficiently with that knowledge.

• Open Psych has struggled to convince AI researchers of the importance of using their expressive representation language and has found that many researchers make do with a simplified version of their system.