Heading Towards the Data Singularity
By Anjali Arora, Chief Product Officer, Rocket Software
Those of you familiar with futurist and author Ray Kurzweil have undoubtedly heard of what he describes as the Singularity — the convergence of man and machine, a seminal turning point at which robots become sentient and can interact and respond to human beings on an equal level. While this topic is widely debated among scientists and philosophers, there is actually another singularity which we are almost certainly headed for. Based on the current degree of integration between data, machine learning and AI, we are in the very early stages of the Data Singularity.
Until recently, these disciplines were pretty much distinct and disconnected. Enterprise data in the last century ranged from executive memos passed around in manila folders to spreadsheets that used VisiCalc or Lotus 1–2–3 to capture numbers related to how a business operated. In corporate environments in the 20th century — what might be called the era of “Small Data” — there was much less information being created about business operations, and there were far fewer ways to effectively capture or manipulate it.
AI was similarly removed from practical concerns until relatively recently. Prior to the workshop that John McCarthy ran at Dartmouth College in the summer of 1956, the term “artificial intelligence” had mostly been used in science fiction books or comics. But even after that formative session, a lack of robust software and hardware relegated AI to the dusty halls of academia, where it essentially languished until just a few years ago. Fast forward to 2018 and we now have machines that can beat humans at poker and Go and are able to teach themselves to play chess at a Grand Master level in four hours.
It’s a different world today for sure, but one with exciting implications for innovation in business. As the data tsunami continues to expand and wash over us, the fields of AI and machine learning are concurrently growing in power and capability. Current thinking is that we generate an average of 16.3ZB of data every year. But the data deluge only continues to increase. Your clothing is sending out signals. Your car is capturing data about the status of other vehicles. Not only are we generating information on our smartphones and laptops, but the burgeoning Internet of Things (IoT) also continues to reduce the viscosity of data. With sensors and actuators appearing on everything from your shoes to your appliances, the amount of information we generate is going to continue to expand exponentially.
As machine learning capabilities get more and more powerful and the rate and pace at which information is generated accelerates, the impending convergence of AI, data science and machine learning offers many exciting possibilities for every business. I see the Data Singularity as providing a tremendous opportunity for companies ready to step up and take advantage of this new model. Imagine being able to easily connect data from your supply chain to market analysis data and using it make strategic decisions about product development. Every day, more and more applications that connect all kinds of business data to AI-driven analytics are coming online, bringing us ever-closer to that being a reality.
The most exciting thing about this convergence is that it has tremendous and broad implications for innovation in business. Smart companies are leveraging the potential to enable informed, real-time, strategic decisions to drive business success. The decisions being facilitated range widely. How do we run our business? How we grow our business? Who do we partner with? Who do we sell to? Who we buy from?
I encourage you to exploit this exciting and disruptive intersection. Here’s some general guidance to keep in mind when exploring the power of the Data Singularity:
· Start small.
· Identify a business issue related to data capture and analytics.
· Conduct due diligence on available AI and machine learning solutions: They range from niche capabilities like image processing to large tool libraries.
· Use the data you have. Rent some from FigureEight to fill it out.
· Build and test an AI instance/machine learning model based on business objectives.
· Iterate!
You don’t have to sign up for IBM Watson to take advantage of the Data Singularity, but you should absolutely get started now. There are already all kinds of solutions available, and more appear every day. Again, they range from AI instances that deliver comparatively simple and niche functionality like image recognition or speech synthesis to applications that use cloud-based APIs and your data to deliver powerful AI-driven analytics. If you are feeling ambitious, and have access to large datasets and an army of tech resources, you can explore more robust options like Google’s TensorFlow.
These are exciting times for business. A new and powerful convergence is underway at the intersection of data, machine learning and AI. Organizations can now capture, manage, organize and rationalize tremendous amounts of information in entirely new ways. This new approach is going to provide tremendous competitive advantages for companies willing to boldly rethink their business models and invest in resources and leading-edge technologies that take advantage of the Data Singularity. So start exploring: It’s a brave new world, with lots of exciting opportunities. And start building your models now. No one wants to be Blockbuster in a Netflix world.
Anjali Arora is SVP and Chief Product Officer at Rocket Software, a global technology provider specializing in modernization and optimization, where she oversees product strategy and R&D. Prior to Rocket, she spent three years as Global Vice President of Software Engineering for Oracle’s Health Sciences Global Business Unit.
