A Customer Experience Approach to Artificial Intelligence.
It’s almost impossible to visit a tech blog or magazine these days without hearing about how Artificial Intelligence (AI) is going to change our lives and put a generation of workers out of a job.
As someone who has been involved with the field for some time — I earned my masters degree on the subject back in 1999 back when there were almost no AI jobs out there — I have a slightly different opinion of the domain having seen it evolve. Especially as it relates to business and the Customer Experience.
For me, AI has always been nothing more than a tool or automation at best. And it can be realized in one of two ways: Augment or Replace. By this, I mean you either augment the human by giving them some silicon assisted superpowers, or you replace them.
Most unsuccessful adoptions of AI tend to revolve around implementations that seek to single mindedly save money. Interactive Voice Response (IVR) technologies, for example, seek to replace human phone operators. I think we can all agree they do a woefully bad job and sometimes add frustration to the customer experience. Bots that don’t reveal themselves as Bots also fall into this category. That’s not to say that all replacements are bad. Online ordering, when designed correctly around user needs, has been a game changer.
The problem with trying to replace a human completely is that computers are notoriously bad at handling novel cases where something goes wrong or there is a special case. They then fail in an unflattering way known in the field as ungraceful degradation. There are even known chess study positions that are easily playable by humans that chess AIs still fail at, simply because they are completely novel and not included in the training data. The IVRs are really no different to some poor souls crammed into an Indian Call center who have been given a script to follow and have no idea what to do when things go off script. In both cases, the business is transferring their responsibility in a very irresponsible way.
And all it takes is one of these outliers to go wrong and an entire brand can be damaged. And here is the kicker: we’re still not getting this right, even as humans! It seems that since the advent of Social Media we’re getting to hear about (and see) humans mess up more and more in the arenas that require nuanced judgment that is currently beyond computers, like marketing campaigns and customer service. Add that to a world of professional victims that is becoming easier to offend with every passing day, and I’d say our jobs are safe for now! You only have to look at Microsoft’s racist nazi-bot to see that full human replacement is not on the horizon.
Augmentation is a different story. I would even argue that since the invention of the hand-held calculator we’ve been utilizing AI in a positive way. As long as the technology is harnessed correctly with well designed experiences, businesses can operate more efficiently and drive the Customer Experience from the inside out through their employees. After all, less mistakes on an invoice mean less car-crashes downstream. These little things add up. Augmentation doesn’t remove human error but it definitely reduces the odds of it. Thank you, Spell Check.
When a business makes the decision to employ AI in some way, it should always come from a place where value-add to customers is paramount. Heck, it might also save some time and money as well, but those factors should not be the primary drivers. Starting with the wrong lens can skew your perspective and may end up costing you more in the long run. By designing our AI through a customer-centric lens, we can set ourselves and our customers up for success by designing experiences that will be reliable, efficient and consistent.