Blog: AI and My Trust Issues

Artificial intelligence. You know no one in marketing or PR came up with that term. It sure doesn’t draw me in — I don’t even like to think about artificial sweetener. But I find myself interacting more and more with AI — some good times, some not so much.

I was among the first wave of Alexa adopters, jumping onto the waiting list of consumers itching to bring a voice bot into my home. Things went well enough with my first Echo that when I upgraded a Sonos speaker I made sure I got one with Alexa built in, too. I ping Alexa daily (she answers to “Echo” on the living room speaker to avoid a cacophony of responses) for the weather and to set timers — and often to have her play music or the radio. When I’m really lazy, I ask her the time.

I have a Google Home, too. At least Alexa has a human name. I really don’t like having to say, “Hey” and then the name of a multinational tech company just to set a kitchen timer or listen to the news while making dinner. And then having to say it again to raise the volume. That Google named its voice engine “Google Assistant” doesn’t make me warm to her, either. It’s like Google doesn’t want us to be friends. At least you can just say, “Stop,” to turn off a timer; it was the bot that gave me that friendly tip.

Siri and I have never gotten along. She never seems to have an answer for what I want to know. And if multiple people have an iPhone, that’s trouble. Last night my partner asked her Siri how much time was left on a timer; my Siri answered: “The timer is not running.” Then Liz had to pick up her phone, get facially approved (great use of AI!) and then tap the clock icon. An analog timer would have been more gratifying.

I use Alexa every day, but I admit we haven’t progressed much over five years. That’s not for the lack of effort from Alexa developers. They seem hard at work coming up with new ways to make Alexa more capable, and recently I’ve noticed her extra effort when I ask for the forecast: “Would you like to know the weather for tomorrow?”

But I have trust issues. Even before publicized reports of Alexa spewing a private conversation to an Echo owners’ friends, I was wary of a smart speaker listening to everything I say so I didn’t give her access to my contacts. Voice engine companies tell us the mics are only triggered by certain wake words, but I know better. I wish I had a dollar for every time Alexa or Siri barged in on a conversation uninvited. Sometimes Alexa wakes up, unsummoned, and says something totally out of the blue.

Eerily, this morning I told Liz my topic for the week was “AI and my trust issues,” and the Sonos speaker chimed in. Nothing in the preceding sentence sounded remotely like “Alexa.”

Consumer-facing companies had already been moving to AI bots for customer service, and they’ve ramped up those efforts during the coronavirus pandemic. That makes sense with call centers affected by new workplace rules. Banks and internet service providers are referencing longer wait times than usual and steering customers toward apps and chatbots. I’m all for efficiency — I hate waiting on hold on the phone — but bots haven’t had the tools to handle my specific needs.

After a fraudulent charge was made on my debit card a few weeks ago, the bank’s AI detectives nailed it when it was highly unlikely that during New York on Pause I had made a $100 charge at a New Jersey gas station at 6 a.m. on a Saturday morning. AI caught the fraud; a real person called me and put me through the paces on the phone to be sure I wasn’t the scammer. All good.

A few days later, my bank statement had three $100 charges, same gas station. The bank warned on the website that hold times to speak to a live person were “longer than usual” during the COVID-19 crisis. But when I tried to use the mobile app or website bot, there wasn’t an option to question fraudulent debit card charges. I picked up the phone and an hour and six minutes later, a live person reassured me the issue would be resolved. I look forward to when the bots can handle more serious issues like mine … quickly.

I tried to go through the Verizon website last weekend to cancel Showtime after the free trial ended. “We’re here and ready 24/7,” the website trumpeted, pushing online services and apps. When I clicked through to the add or change service option, I got a “Bad Message 431.” I picked up the phone, and just under 10 minutes later, I was speaking to a live person, whose job was to pitch me on router protection before processing my cancellation.

I know AI is making day-to-day life better. If you had told me 10 years ago I’d be depositing a check by phone someday, I’d have been amazed. With a taste of what AI can do, I want it to do it all — and better.

The post Blog: AI and My Trust Issues appeared first on EE Times Asia.

from EE Times Asia https://ift.tt/2XfkMzp