Articles
WINTER 2015 71F rom an industry perspective, the era of cognitive computing has dawned with the promise of human-centered cognitive prostheses, as just one of many benefits anticipated (Kelly and Hamm 2013). Gartner, the technology industry analyst firm, projects that by 2017, 10 percent of all computers will be learning (Plummer 2013). However, scholars familiar with the field of artificial intelligence and cognitive science are rightfully cautious, after witnessing a roller coaster of ups and downs over AI's relatively short 60-year history. Simply put, hard problems still remain.So how can we know if this time is really different? After all, three of the key pillars of cognitive computing, namely machine learning (ML), natural language processing (NLP), and hypotheses generation with evidence-based explanation (EBE) capabilities, have existed for quite some time. The first author personally recalls programming hidden Markov model (HHM) learning algorithms for speech recognition in a