Hacker News new | ask | show | jobs
by mastef 4117 days ago
Do you have any specific ideas of a massive road-block that would benefit from a Mechanical Turk approach? E.g. issues that would require an AI, but if split into a lot of tiny problems, that could be solved as a whole?
1 comments

I've been giving this some thought and have some ideas on a few directions we can go in. For example, cognitive biomimicry/AI can be trained to predict certain things like answering "What is the capitol of Austria?"

With statistical inference and probability, a system can provide a number of answers to questions. A crowdsourcing component can be used to rate the answers the system provides thereby allowing the system to learn via feedback loops.

Apply this to Life Sciences, molecular biology, genomics/proteomics and we may have a system intelligent enough to produce new hypothesis in the area of life extension research or a system intelligent enough to combine two pieces of knowledge to come up with something new, a discovery or series of discoveries.

I have some other ideas too, look me up at biomimic@gmail.com if you want to collaborate on these.