The Future Of AI: Killing On Hold Music

(Read Part 1, Part 2, and Part 3 of the Future of AI series)

Back in 2016, I went out on a limb and published an article that said the call center will be dead by 2019. I was so confident I even transcribed the article into Mandarin. Boy was I wrong! And the comments said before and after 2019 ridiculed my prediction - see them here.

Now I make my next prediction: the only service that AI will make irrelevant is hold music, also known as music on hold (MOH).

Everyone of us knows the business practice that plays recorded music or speech to callers who are on hold. The purpose of hold music is to prevent callers from thinking the line is dead and hanging up. It can also help to calm and distract callers during stressful calls.

But my new prediction is that by the end of 2025, your call will be answered on the first ring and you will not be offered numbered menu items to direct your call.

In fact, you can create your own AI bots (optional) to have conversations with call centers that instantly schedule appointments, gather healthcare test results, or reduce your mobile phone bill.

But first lets start with the weakest link: humans not believing that AI cares.

Robots to the Rescue

Before robots can help us, we have to accept them.

In Knoxville, Fai Thai Kitchen in Bearden Village has a robot that delivers its food inside the restaurant. Fai Thai is not getting rid of human jobs. It's just cheaper to use a robot to deliver food to the tables.

The restaurant is small but it has a lot of customer traffic. Fai Thai Kitchen only has three or four employees which consist of Kultida "Fai" Blais, the wife, her husband, and Fai's Thai grandmother who are sometimes too busy taking orders and actually cooking the food to serve the customers.

Rosie the Robot delivers food quickly and efficiently. But it also allows when things slow down, for Fai, her husband or the grandmother to come out and talk to customers. Get to know them. Build relationships. And allows customers to take selfies with Rosie, affectionately block Rosie the Robot from getting to its location as a joke to see its cute reaction, and watch kids run along Rosie as she returns to the kitchen.

Rosie the robot does exactly what AI needs: brings emotional intelligence or EQ to its task. It creates empathy within the humans that need Rosie to do what she needs to do.

AI Needs EQ

Empathy in humans is real.

Empathy within AI is "faked" or to be manufactured. I published an article during the dark times of COVID in 2020 of how AI requires EQ and how it can be generated to gain acceptance by human beings.

Examples of AI faking empathy is by using augmenting weather data in preparing responses such as "Wow! Isn't it cold outside?" Or "I hope we get some rain soon." Little banter that allows the human to know that the AI is in the present moment and they understanding via geolocation where you are.

More importantly, its not just about offering an open ended question and then not following up, but listening to the human response, "No, I love cold weather!" Or, "I live for sunny days. Do we really need it to rain?" And the AI returning another volley of emotional intelligent statements to create rapport.

Reinventing the Psychology of Zero

Since the introduction of automated telephone calls: "Press 1 for English", "press 3 for customer service", humans ignore all of that and keep hitting the zero button until they are connected to a living, breathing human being. I call this the "psychology of zero".

Humans have no patience with automated system that get them stuck in a loop while listening to hold music. Zero, zero, and zero to get to an operator.

What if we got rid of the automated menu and the hold times, and instantly connected you with a cheery voice to ask how they could help?

Note due to regulation (which is warranted) AI has to state upfront in order not to mislead that they are AI. But to counter and prevent "psychology of zero" AI has less than 0.001 seconds or basically zero to come out with a witty remark, empathetic statement to keep the human "hooked" to the call.

Now the "psychology of zero" is simply a race to say the most endearing thing to engage in a human way.

The New Human+AI Way

With new models such as Google's NotebookLM or Meta's open source version NotebookLlama, are creating engaging banter and statements to keep the human rooted in the conversation is as easy as generating a fake AI podcast.

Listen to some fake AI podcasts based on articles and interviews I have done using Google's NotebookLM. Here's one talking about my discussions and articles around AI regulation. And here's another where AI created a podcast around strategies based on an article I wrote around the 7Rs of cloud migration.

Can you hear the pauses for breath? Can you listen to the AI podcast hosts interrupting each other, making analogies, and even attempting humor?

This is the new human way for AI. Telling stories in a way that is engaging. Using the words that it has been trained on to create a beginning, middle, and end of explaining something.

This feature of creating AI podcasts requires human and AI to work together.

Humans empathy collected together in banter and statements that can be recycled.

And as I led with my prediction that by the end of 2025, hold music will be dead - it will require a rethink of how call centers work.

One human call center agent to handle one hundred AI led calls. But at anytime, when the human feels that they are not being heard or understood. The human can simply say, "You are not understanding me." And the system triggers AI to redirect immediately to a human that has the context of your query on their screen to answer your issue.

Sure, people might then when they hear they are talking to AI simply game the system by saying "You don't understand me." But the goal is to create a Rosie the Robot situation where it would be fun to joke with a robot so why not?

To Bot Or Not To Bot That is the Question

The call center system also needs to consider the next generation and not just Boomers and Gen Xers. Digital natives post-COVID are still having issues talking to humans or making phone calls. Millennials and Gen Z are facing challenges with in-person interactions, primarily due to the significant time they spent communicating online during the pandemic, leading to potential difficulties with social cues, building relationships, and even navigating office dynamics.

So it would allow for bot to bot interactions and allowing your phone to engage with a call center to share information that instantly schedule appointments, gather healthcare test results, or reduce your mobile phone bill.

The future of call center systems allow for bots to recognize bots by using a high pitch tone at the start of a phone call to override voice models and go direct with API push and pulls.

Then if its not bot to bot, allows for a human to interact with a voice to voice model.

An AI voice-to-voice model is a type of artificial intelligence (AI) that can convert spoken words into different voices. This technology is also known as speech-to-speech technology.? It allows for more efficient interactions as its cheaper infrastructure costs as there's not a constant translation from voice to text back to voice. Vocal cues and statements are locked into the model.

AI voice-to-voice models use deep learning algorithms to analyze and learn from human speech. The AI then uses this information to generate a vocal model that can replicate the original voice's tone, nuances, and articulation.

Then again, when the human voice repeats the same question twice, three times, or raises their voice or states, "You do not understand me" the system triggers out to a human agent immediately.

What gets killed in this scenario? On hold music.

AI has to become more Rosie the Robot. It has to serve a task that allows for more humanizing. Slowing down to get to know the customer even though technically its speeding things up (no hold music, no waiting in line, and offering direct API calls).

The fastest way to a customers heart is being efficient at solving its problems.

What thrives in this scenario? AI that is learning empathy and contextualization relationships with human clients.

And what's more human than human? AI learning to care.