Why ‘augmented intelligence’ is a better way to describe AI

There’s a lot of talk about artificial intelligence, and to be honest, most of it is hype. In fact, a scary part of the AI startup ecosystem is that there’s no regulatory board that’s actually kicking the tires of different companies to verify that their technology is AI. To add insult to injury, the term artificial intelligence is a misleading label for the methods behind AI – machine learning, deep learning, and natural language processing. In their current state, these technologies are much better at helping professionals improve their performance than they are equipped to replicate human intelligence with an artificial entity.

The term artificial intelligence is a misplaced signpost that doesn’t get us where we need to go – especially when applied to highly human sectors, like sales and customer support. Artificial implies that the data scientists and companies deploying these technologies are replacing humans with machines. This is an incorrect assumption. Scholars and industry experts agree that human-level artificial intelligence is years away. A recent survey published in the MIT Technology Review estimate another 15 years until AI can imitate a retail salesperson, and 120 years until the automation of human labour.

A better way to think of artificial intelligence, is in how it’s leveraged to augment human intelligence. One example is in using AI to improve chat agent performance. It’s the best application of the technology now. In its current implementation, AI is capable of two things: (1) automating repetitive tasks by predicting outcomes on data that has been labeled by human beings, and (2) enhance human decision-making by feeding problems to algorithms developed by humans. Our live chat agent training solutions help the humans behind successful sales and customer support interactions perform better at their jobs. We do this by employing predictive outcomes based on unique data and customers. Our augmented intelligence helps chat coaches develop unbiased, effective training strategies to empower live chat agent performance and achieve significant operational gains.

The emotional intelligence of humans is what drives sales. EQ is a human competence that’s difficult to teach machines because it involves intuition, empathy, and moral judgment. These traits make up the qualitative aspects of human interaction that are difficult to quantify and draw insights from using data. By testing conversation variables across hundreds of dimensions, we’re able to reverse engineer a conversation to give EQ recommendations to the human agents that can carry them out. When applied to live chat conversations, augmented intelligence improves live chat agent performance and happiness by optimizing the task at hand.

Artificial intelligence – and chatbots – can’t convince people that they’re human in conversation. We took it upon ourselves to face-off with award-winning bots and found that customer support chatbots fail the majority of the time. In our recent work with a client, we measured the optimal chat agent response time in relation to conversion rates. We found that if the live chat agent responded too quickly, the customer was less likely to buy. On the other hand, if the live chat agent responded too slowly, the customer perceived a lack of adequate customer service.

This goes to show that automating interactions with customers through adolescent technologies like chatbots isn’t going to convert site visitors, let alone increase customer retention. There’s an optimal behavior for each aspect of a brand to customer communication, and it changes on a case-by-case basis. The bottom line is don’t let a chatbot be your brand ambassador. Why? Because they’re unable to sell effectively, they don’t reflect brand identity, and they’re not personalized to adapt to the user.

Leave a Reply

Your email address will not be published. Required fields are marked *