In data we trust: the Dataist Company

For years we have been listening to famous quotes, almost turned into mantras, related to customer centricity. Companies need to listen to their customers because, well, nobody knows a customer better than himself, right?

According to dataism, that is no longer going to be true. Let's use the definition given by Yuval Noah Harari in his book, Homo Deus:

"Dataism declares that the universe consists of data flows, and the value of any phenomenon or entity is determined by its contribution to data processing."

Dataists claim that humans cannot deal with current data flows and for that reason they cannot distill data into information or even more, into valuable knowledge. Therefore, dataists are skeptical regarding human knowledge, and they prefer to put their trust in big data, algorithms and machine learning. So, it is time for companies to become dataist?

In the same book mentioned before (Homo Deus), professor Harari also explores what he thinks makes a data processing system improve (actually, dataists think that we can interpret mankind as a data processing system in which individuals act as chips). So, from a dataist point of view, companies are able to improve through the following:

• Increasing the number of processors: we are already seeing that in the AI era, data is going to be the new competitive advantage. Companies need data to train their algorithms. The more data, the better the algorithm. The better the algorithm, the smarter and more efficient our customer engagement and internal processes get.
• Increasing the variety of processors: mathematical algorithms create value out of data. Companies able to apply a large variety of algorithms and AI to deal with different types of data (from structured and non-structured databases to speech and video recognition) will be the ones in the best position to compete in increasingly competitive markets.
• Increasing the connections between processors: this one might seem obvious and complements the above statement, but companies can only get the most out of data if not only their datasources are connected, but the processes managing it. Our customer engagement will clearly improve if we can connect the data and algorithms dealing with natural language processing (NLP) and voice and video sentiment analysis.
• Increasing the transfer of information between connections and processors: What it is not as obvious (or is it?) is that companies connecting the same data through different internal processes can multiply the benefits.

For example, imagine that I am a customer of Amazon contacting their support because the shipping company has already lost my order during its delivery route (true story). Well, Amazon can process my voice tone and what I tell the contact center guy and conclude that I am angry (I really was). Therefore, Amazon offers me to send another replacement and to extend my Prime membership for a month for free which is good. But guess what would have been even better. Connecting with their internal processes and telling them NOT to send me the second replacement through the same carrier. Yes, mr. Bezos, you needed to send the same order three times through the same shipping company for me to receive it once. Not a very lucrative business, sorry.

After reading all of the above, it might seem that the human factor that has been praised all over the way as the real business changer of Digital Transformation is not going to be important anymore. Well, I firmly think that there is still room to be optimistic:

• Even if we think that the singularity will arrive and AI will surpass humans and be able to improve itself autonomously, we are still in charge of creating the connections and processors dataists trust. Fortunately, this means (at least for the moment) not only programming algorithms, but understanding the needs and feelings of customers and internal teams. In the example before, it still gets a human (in this case, myself) to detect that I was really pissed when I saw my order going through the same shipping company for the third time before changing that particular process. The more multidisciplinary the team working with the data (from linguists to interaction experts, and of course data scientists) the more powerful the solutions that we will create.

Now the bad news:

• How long will it take for algorithms to detect those kind of situations and even more complex ones? There is a lot of books written around this topic.When the day arrives in which a general AI surpasses humans, will our value as a species be reduced to the amount of useful data we are able to generate?