I was involved in the development of AI applications for a large company. A dozen people developed applications for five years. A few applications were very successful, but most failed to be what was expected. They were not used, were not trusted, or proved to be far too expensive to maintain.
As the group faded away the outgoing explanation was that the artificial intelligence was still there, but was now embedded in applications. Exactly what it was, other than complex computer coding, was less well defined. The default answer seemed to be that the AI elements were not doing human style reasoning, but were too complex for normal humans to fully understand. A strange definition by itself.
An article in Wired brings up this same point. What is AI? It positions it as clever methods, but not necessarily the ones that humans would use to address the same problems. A number of interesting examples are given, including a bot warehouse at diapers.com
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment