For every great advance in artificial intelligence (AI), offering a chance for machines to truly think and be more useful than ever before, there is always a negative story supporting the fear that humans may one day be subservient to their creations. A recent online beauty pageant with 600,000 entrants using an AI judging process was blasted by critics when almost all the winners were white.
Of course the AI system needs to learn from how someone defines beauty before it can start teaching itself from additional examples. In this case the vast majority of images that were initially used to define beauty were images of white people. Skin tone was not a factor in the judging process, but with the desirable attributes mostly defined as those more common in white people it’s hardly a surprise that the AI system effectively learned how to be racist.
AI is already entering use in ways that many people may not yet have noticed in their normal daily life. When you call up your mobile phone provider and an automated voice says that you can ask a question using natural language, rather than going through a long menu of choices, then you may already have run into the commercial use of AI in the real world.
Companies are finding that customers are giving up much more information on their preferences and spending patterns than ever before. Smart companies are connecting together loyalty with payment systems and creating a new level of service so it’s less desirable to do things the old way. Starbucks is a great example. If you use the app to order and pay for a coffee then you don’t need to wait in line because it will be on the counter waiting for you to pick up – you can grab the cup and leave immediately. That’s better for the customer, but it also means that Starbucks builds a database of everything you have ever ordered, where you go, what time, and how often.
Now combine predictive analytics with this data and the company can intelligently predict what the customer might want next. At the very least it allows the company to send very specific recommendations for new products, or offers, that are targeted at one individual customer. With location awareness, retailers can even send time-limited offers to customers that are in-store, nudging them to make a purchase.
Combine this knowledge with the natural language speech analysis that is becoming common in many contact centres and there is a powerful new level of engagement between companies and their customers emerging that is being driven entirely by AI practices.
On the whole this greater intimacy is better for customers and brands. The companies that are able to analyse their customers in this way will be able to build a far better experience that drives greater loyalty. The customers should find that the companies they interact with regularly know exactly what they want – almost before they ask for it.
The only fly in the ointment is that this type of business environment relies on companies managing far more data on customers than ever before – and for a fully intelligent system to work, it needs quite intimate data describing payments and preferences.
I believe that some companies will get it right. Look at Facebook for example. Customers hand over extremely private personal data everyday to a social media giant that then sells ads around those customer preferences. The users accept that they need to supply their data and in return they get a useful (and free) networking service.
The Starbucks app similarly offers a real advantage over the old way of doing things. Who would line up and wait for a barista to make a coffee when you can order from a bus five minutes before arriving at the coffee store?
But if these real advantages are not present then I think companies will struggle to get customers to hand over enough personal data for them to be able to build truly intelligent systems that can predict customer behaviour and needs. Offering a few points on a loyalty card doesn’t cut it today in an environment where customers see leaks of personal data making front page news stories.
AI is going to be a major force in changing the way that companies relate to their customers in the coming years, but this challenge of encouraging customers to willingly trade in their privacy in return for a better service needs to be addressed head on or the full potential of these technologies will never be reached.
BLOG: Mark Hillary