isn't it fun how quality standards just do not apply to AI?
I just had to call a customer service hotline - but now they put you through to an AI. I had a pretty simple request, a simple piece of information to be relayed to a responsible party. But the AI couldn't do that, so we went in circles (drip-drop goes the water~) until I finally got put through to an actual human person employee, who took my message and e-mailed the person and that was it.
BUT what's interesting is: I got the announcement "Your call might be recorded, please stay in the line to evaluate this conversation-" between my "conversation" with the AI and before I got put through to the human employee. Not before talking to the AI. Before talking to the actual person. Mind you, the AI failed this pretty simple task and went in circles and was useless. The employee was friendly and immediately helped me, no problem. But I can only evaluate her, not the crap AI.
Any negative evaluation on my part would reflect poorly on the person who helped me, so obviously I give a positive evaluation - but that also means that from the perspective of the company, customer satisfaction remains constant ("yay AI was a success!") OR it gets worse for the human employees when people take out their frustration with the AI on them. Pisses me off.
