VoiceSage’s Matt Weil argues that naive AI coverage harms our appreciation of what this useful technology can actually do
In the summer, a story gripped the world’s media: a sinister-sounding AI alert.
The claim was that Facebook had had to end an AI-powered chatbot trial that was deemed a threat. And as a result, the press had a field day: ‘Robot intelligence is dangerous: experts’ warning after Facebook AI develop their own language’ was typical of the headlines the scare generated.
The story was about the fact that two chatbots had created their own ‘secret language’ (doubtless to be used for that Robot Takeover). However, all this stemmed from an innocuous trial of Natural Language processing in the context of negotiation. Facebook AI coders had been experimenting with programs that negotiated with each other as part of a project to understand how linguistics played a role in such discussions: the software were programmed to experiment with language in order to see how that affected their dominance in the negotiation, and they worked out some clever shortcuts.
We soon witnessed stories that corrected the misleading impression that there was anything sinister afoot: as Gizmodo says, “In their attempts to learn from each other, the bots thus began chatting back and forth in a derived shorthand. But while it might look creepy, that’s all it was.”
However, it’s adding up to more fuel for public unease about AI that is misplaced. And of course, too many people don’t understand AI in general. That is a problem, as what chatbots (the most immediate and eminently practical application of AI ideas to our space) can and can’t do really matters in a business and customer service context.
Over-hyping has negative repercussions. Early usage of chatbots has in many cases disappointed brands as they are pushing them to replace humans instead of augmenting them. Take this recent discussion, for example, Putting Human Call Center Agents on Hold . It’s a well-researched overview, but it’s predicated on the claim that when it comes to AI in customer outreach, it’s going to be all about replacing agents – “The impact of artificial intelligence on contact centre operations will be significant… Chatbots are the first step, but automated bots that naturally interact with callers will initially reduce and eventually eliminate all but the most complex human agent interactions during the next decade.”
Although this author admits this probably won’t happen inside of five years, he’s certain that it’s inevitable.
[easy-tweet tweet=”Great brands like Twitter and Airbnb are already successfully using chatbots” hashtags=”AI, Chatbots”]
Great brands like Twitter and Airbnb are already successfully using chatbots. And there’s no question that chatbots are a really useful technology, but the fatal error is to think chatbots are all that is required.
The public likes problems solved quickly, and a slick interface that gets issues like updating an address completed without having to speak to a human after multiple levels of ID check. As a recent Business Insider analysis shows, 44% of US consumers said that if a company could get the experience right, they would prefer to use a chatbot or automated experience for CRM.
Chatbots are not a panacea
But note the key phrase – if a company could get the experience right. Customers very good at spotting over-eager AIs. If a chatbot fails to help you find out if your delivery will come in on time, the experience becomes a negative one. Brands pushing too hard to automate too much of the customer outreach process – seeking cost-savings and efficiency – will face a backlash as a result.
Brands are correct to value the concept of automating some of the outreach process, and the efficiency increases that follow due to agents being moved to higher value work.
However, just as we need to exercise some scepticism about headlines about a robot takeover tomorrow, organisaton also need to exercise some balance about the right use of AI in customer service.
Chatbots are useful for firms, but be realistic – they will work as an addition and a support for the contact centre, not a replacement. Use AI/chatbots to automate simple, repetitive tasks the user needs to do, or as a fallback, but it would be foolish to adopt them too quickly, nor make them the only arrow in your proactive customer service quiver.