Artificial Learning and Machine Learning for Indian brand engagement? Gimme humans, please!

• Srinath Sridharan, Corporate Advisor & Independent markets commentator, pens an OpEd on issues that brands create by claiming to use AI / ML, when basic processes don’t work.
• From a consumer perspective, he shares his experience of using one of India’s largest OTT apps and questions whether all customer data can a broadband provider read from the apps that are built-in or linked to its usage.
• He further highlights how human interaction is more valuable for a consumer than AL and ML interaction.

I’m not a techie or a deep-techie. I’m also not a patient consumer, specifically when it comes to tech-related trouble-shooting. I prefer to deal with brands that claim mastery over remote trouble-shooting or offer anything that reduces human intervention or even the remote-requirement for it. Just as long as their own technological prowess can solve my issues that their own brands create for the consumers in the first place! But this reverie got shaken recently.

I have had a subscription to a global OTT player since its launch in India. The app was linked to my secondary email ID. During the lockdown, the platform’s content-creation capability, along with its technical ability to deliver different content on different screens (phone, tablet, TV) simultaneously was proven too.

But that’s where the good part of the story ends!

My TV has been linked with the broadband provided by the largest technologically disruptive brand India (or even the world) has seen in that space. The broadband subscription has been linked to my primary email ID, and the OTT app, built in with that broadband, had been working fine until recently.

The tricky one

Over the last 2 months, I wondered why the app was not allowing me to login from my tablet when the kids were streaming video on that app on my TV. Surprising, considering that I was on the app’s highest subscription slab!

Trying to decipher it, we dialed their call centre and the executive guessed my trouble could be due to the recent RBI diktat around recurring payments on credit cards.

Well, that was not the actual trouble, or the solution! I had to try getting in touch with my broadband’s customer care team. Try and try as I did, through the app, their WhatsApp chat, their acclaimed chatbot, their voice chat as well as the call centre, it took me either on an infinite loop or asked me to the select options that the broadband provider felt consumers might or should have. My issue seemed to be out of their purview. Nothing (has) worked yet.

After talking at length with the OTT brand, I figured that the broadband provider had over-ridden my OTT subscription plan into the basic OTT plan that the broadband provider offered as a freebie. And hence I had to try reaching my broadband provider, to ask someone to solve this issue. Again, I found no luck reaching a human! This experience only created brand dissonance and disappointment.

Probably because AI engines still don’t understand the frustration or emotion in the consumer’s voice, they miss the critical part of the consumer-grievance-process: let the consumer vent, let us acknowledge the trouble and then solve it.

All I wanted was simply to de-tag my OTT subscription (linked to my secondary email) from my broadband. Sadly, I might be one of the few consumers who had such a simple issue that the problem-solving simulations of the giant might not have even thought of.

The larger issue

But then, I’m a simpleton and not an expert in databases or the AI / ML construct. Suddenly I have a new (additional) worry on this persistent issue: How can a broadband provider override a subscription that I have with another service provider (OTT) and override my decision of an OTT subscription plan? Especially when that OTT is not linked to the registered email of the broadband subscriber and when my payment instruction is tagged to my OTT subscription?

How much of the customer data can a broadband provider read from the apps that are built-in or linked to its usage?

AI, Brands, Consumers, Dilemma, Ethics

In 1950, Alan Turing first ideated on the ‘thinking machine’. Since then, Artificial intelligence (AI) has seen a lot of progress.

The field of AI is constantly being advanced with evolving technology and technological computing power. While it may be difficult to exactly pinpoint where AI currently is, it’s far easier to describe what it surely isn't. People think AI is a smart robot that can do things better than a very smart human and that this robot knows it all and could answer any question.

AI refers to computer programs that can process various analyses and use some predefined criteria to make decisions. One of the long-term goals of AI is to have the smooth ability to communicate effectively; currently, the chatbots and machine learning-based language processors still can’t infer meaning from human statements or understand nuance. The most decently capable AI systems are wonderful products that follow processes defined by smart people. The AI systems can’t make decisions on their own!

AI is capable of doing many things that human beings simply cannot, such as calculating enormous sums or sifting through tremendous amounts of data. It is a fair worry that many have with the notion of what role humans would have, once these thinking machines become capable of operating independently.

In machine learning (ML), which includes deep learning and neural networks, an algorithm is presented with lots of ‘training data’, the examples that the algorithm learns from. The important aspect is that it is still categorised by humans, until ML can complete the task on its own. The best-in-class ML algorithms are learning from their machine memory and using and executing statistical models. Generally, when (most) humans (me included) think about AI, they are confusing it with ML!

Brands are getting carried away in their exuberance to showcase their ability to use financial resources to invest in technology to solve what they think as ‘all of consumer issues’. That’s the first fallacy. Sadly, the consumer falls off between the two stools - technology and consumer-care processes!

There is a lesson here for marketers: Program your processes for robustness, as if the consumer is dumb and does not know how to use it. But treat the consumer, including the irate ones, as intelligent people. Consumer vanity is the actual reality for the cash box ! After all, you (the marketer) have to make a decision - Do dumb customers buy your products or the intelligent ones? Choose carefully, for your consumer is watching and listening!

For me, the moral of this unsolved OTT-Broadband chase-story is that, until there is substantial improvement in AI / Ml and it’s responsible adoption by brands, human interaction still would be needed to actually solve the problems, created either by them or those machines in the first place!

To quote one of the most astute consumer activists ever in the history of product development and disruptive-consumption, Steve Jobs, “Start with the customer experience and work back to the technology.”

The worries about how AI will be used by brands to impact consumers, and the dilemma about ethics and how society shapes its acceptance and the usage of AI is open for debate. Till then, the only AI/ML I should probably stick to is the local Marathi phrase - “Aai, mi kay sangu?”