- Announcement about May 10th, 2018 update and your password.
How can I start a big volunteer software project like Linux was for intelligence that can end human work? A humanoid robot can physically do jobs designed for humans but without great software it's only a large RC toy or a single task repeat performer. I can finish my robot but to get truly intelligent software that can learn K-12 and beyond needs big help.
Having studied AI on the back burner since high school, I learned the Cyc project was based on the Columbia desktop encyclopedia and after reading a few entries I see that explains why Cyc has 4th grade performance - the source material is written at a low grade level. I looked up a few words I'm familiar with like Jesus and church finding the entries are not up to the standard of an average pastor let alone the best. Also, a Watson video says it can't understand Bible parables. So I propose starting with the top level intelligence which the current best projects can't do, the Bible, and the best published interpretation I know of, the Collected Works of WL by LSM.org as a start then advance from there to things like contrasting Jesus and Paul. If software can achieve that level, not by flat neural nets which cannot answer interactive questions but by expert If-Then-Else training so it can converse, it should be easy to train in lower fields until there is enough coverage to make life easier.
Being 63 I'm starting to feel what it's like to be old and I've also communicated with homeless and 3rd world people who have a great need for all of our technology in a can. It would improve care in retirement homes, group homes, and all homes doing chores that humans don't do. What will happen when you grow old and you have no one to take care of you? I think this project could help everyone. But how can I get volunteers? Thank you.
PS: I have started a new ruletalk computer language which will make expert system methods easier. The parser works for simple input and the vm requires much work. Currently I'm building it on Pharo Smalltalk but have been considering conversion to Python. It could also be converted to Spin to work on Propeller chips with large memory access, or any hardware is possible. But it needs audio, video, and USB input and output (A/V/U). The FORTH people have their own multi-core chip I'm not familiar with. Any suggestions to get the best price/performance?