If you program your learning algorithm to “solve” customer problems in the shortest amount of time possible with the least amount of concessions possible, it will act exactly as I just described. The company would have to be run by buffoons to give the phone machines the ability to change user account information or have the ability to issue refunds, so the end result is that they can only answer simple questions until the person on the other end gives up.
echodot@feddit.uk 11 months ago
That is not how AI works.
It’s not programmed at all it’s a developed network it evolves in the same way that the human brain evolves saying it will try and solve the problem in the shortest possible time is like saying that human agents will try and solve the problem in the shortest possible time. It’s a recursive argument.
And you might say oh but an artificial intelligence could never possibly match the intelligence of human but why would that be the case? There’s nothing magical or special about human intelligence.
doctorcrimson@lemmy.today 11 months ago
Wow you really went off on an irrelevant tirade, there. There is a defined accuracy when you set up the learning algorithm, there is an end goal result that you define with which the program chooses and eliminates “choices” for a given generation. You program it, it doesn’t magically conjure from a witches cauldron or a wish from a genie. And also, we’re not talking about actual intelligence and sentience here, we’re talking about AI as in modern Learning Algorithms, as I explicitly stated at the start of this thread before you used the term AI for the first time in this thread.
With your repetitive, nonsensical, baseless logic I think you would pass for one of those glorified chatbots.