Were I a young man, I would be rushing to develop an AI app which could give me a quick evaluation of how sane some comment or question is. Not, mind you, “truth”, but just common sense-wise.
Ideally it would only be trained on academically accepted reliable sources, like Wikipedia or encyclopedia Britannica. Then it could be fed statements, or questions or whatever and would respond with a common sense analysis.
At one level, it would simply give a few words. “I don’t know”, “impossible”, “unlikely”, “maybe”, “almost certainly”, “definitely”. Just like your mom.
At the next level perhaps a paragraph or two explaining why it reached that conclusion. And finally, perhaps, a reference to the sources it used.
I’d want it to answer complex questions “can you walk barefoot on burning coals without being burned”, “will a guilty person sink in water.” As well as stupid ones “can you live forever by eating jade”, “why did the chicken cross the road.”
Could it do it, I think. Common sense is basically weighted logic, contradiction, and unknowables factored into reliable axioms. So why not?
I’d sure like to see someone try as soon as possible. Common sense seems to be in short supply among the current batch of RI (real intelligence) humans.
HuntingtonHarborThoughts 