Abundance is not Wealth: from Big to Smart Data

We have spent the past few years in sectors where humans are painstakingly trying to make platforms and machines smarter. AI or Artificial Intelligence as we call it. However, something we often end up discussing is what benchmarks should one use to measure intelligence? Is it essential for machines to pass the Turing test or should we selectively identify areas of intelligence where we know machines can perform way better than humans, such as complex problems in resource optimization?

What we have realized and want to share is, it is not a constant stream of data that makes a device or platform smart. In fact, even the best Companies out there probably have little clue of what to do with Tb and Pb of data they collect everyday, every minute, through to every second. Rather what is needed is collecting meaningful data with a pre-defined purpose as part of the overall strategy that can then be used (via machine learning) to automate and make online platforms and connected devices truly smart.

The problem that we face today is everyone is collecting data and yet very few have a proper ‘business’ case in place to loop this data/learning back into the product, monetization plans, revenue modeling, marketing strategy, etc. Sadly, without understanding and applying old fashioned statistics and data mining methods in first filtering the humongous datasets, machine learning is rendered powerless as it tries to look for signal in the jungle of noise.

Leveraging the power of data requires focus. The aptitude of data scientists to identify patterns (through feature engineering, unsupervised and supervised machine learning, etc.) is wasted without the ability to relate these patterns to user behaviors, device behaviors, among others, which requires profound domain knowledge. What is limiting data-driven growth across industries today is not access to the data technology (Big Data, clusters, data warehouses, algorithms and deep dive analytics) but connecting this technology to proper use cases, which will impact the business at a fundamental level! Embedding data scientists, data analysts, data engineers and developers in cross-functional Business units is, therefore, super-critical.

Today’s main issue with the buzz around data science, artificial intelligence and big data, is that this technology is still not connected with the day-to-day needs of companies. Before connecting devices and platforms, businesses need to connect their business strategy to data strategy.

So data is indeed the new oil. But didn’t oil exist even before humans? It is our ability to refine this crude oil into a source of power that still makes it invaluable. The same has to happen for data to truly harness its power.

Leave a Reply

Your email address will not be published.