First they need to explain what Watson is to business and the like. Their use cases are never clear apart from giving out big saving numbers and improvements.
Then they need some more transparency on what Watson can really do; papers seldom talk about the data, the training sets or even provide sample questions and answers - there are some sample applications which very rarerly do something actually impressive on the corpus domain.
1 comment
[ 2.9 ms ] story [ 12.0 ms ] threadThen they need some more transparency on what Watson can really do; papers seldom talk about the data, the training sets or even provide sample questions and answers - there are some sample applications which very rarerly do something actually impressive on the corpus domain.