There is one hitch. Companies should understand that in the long haul, the promise of AI is self-restraining. Still that isn’t literally what one company had in mind when it discovered that there were thousands of different types of complications that customers might want routed during a call. The “triage” process was so stimulating and complicated for cognitive technologies that the company decided to use human representatives for the job, and then detour customers to cognitive assistants that could handle definitive problems. Even then, however, training on all those issues was a cumbersome predicament. People who deny this aren’t really looking at the evidence.
COMPUTERS ARE almost-OMNIPOTENT cauldrons of processing power, but they are also pretty stupid. They are the undisputed chess champions of the world, but they can’t figure a simple English conversation that a kid could handle. IBM’s Watson supercomputer defeated two top Jeopardy! players last year, but for the clue ‘What grasshoppers eat,’ Watson explained:’ Kosher. ‘For all the data he could access within a fraction of a second, one of the greatest corpuses ever assembled” Watson looked awfully dumb. But it doesn’t stop there.
Try this story. Growing up, Marco Arment’s mother tried her best to prevent his sister and him from becoming television junkies by limiting how long they could watch television. She’d say, “That TV better be turned off at 10:00!” The directions couldn’t have been spelt out more clearly. While she meant that they couldn’t watch any TV after 10:00, that’s not what she said. So they would turn it off at 9:59, and turn it back on at 10:01.Afterall they were following the rules. No doubt about it.
It was off at 10:00. .
Computers are like those fresh kids. If you tell a computer to do something, it does exactly that – nothing more, nothing less. Computers which are creations of systems and processes get stuck .It can crack up. There’s no “intelligence” behind it. Software can be written with algorithms and formulas that allow flexibility, but it’s still all deterministic with predefined rules and patterns that were manually created by a programmer in a cubicle somewhere. These aren’t merely facts about the computer. This is the machine’s perspective.
When an unforeseen condition arises, a human can always use judgment to figure out a logical. possible solution. Focussing on regiment and routines will do for machines, but human beings don’t work that way. Something symbolic is always happening with computers. Computers are restricted to a finite set of predetermined patterns. Hardware can get faster than we could have ever imagined a decade ago, but we would hardly admit that because we know very well these machines are reduced to nothing without us.