‘We are entering a third era of automation, in which machines encroach on decision-making. But there is still a role for wetware.’ - Tom Davenport, Only Humans Need Apply: Winners and Losers in the Age of Smart Machines (2016)

As we work day in and day out using analytics to drive innovation in our organizations and in society, I think it’s essential that we realize our data science today is very much shaping the way we will live and work with machines in the future. That’s why I was very happy to see how Only Humans Need Apply: Winners and Losers in the Age of Smart Machines emphasizes how we need to be conscious of the current artificial intelligence revolution and really work to make sure human intelligence remains complementary.

“Oh fantastic,” I hear you sigh, “Another ‘must-read’ to add to the list. Because I’ve just got so much free time to read these days.” Well, that being the case, I still recommend you make some time for this one during whatever summer holiday you manage to carve out and….watch out…I’m going to add a few more essential whitepapers to your list a bit further on.

From Stephen Hawking to Bill Gates through to Elon Musk, big brains have highlighted both the potential, as well as the risk of Artificial Intelligence. Logical then that Tom Davenport contributes a ruthlessly pragmatic, but still fun to read, account of what we as knowledge workers can do now, not only to keep up (and keep our jobs), but also to ensure that ‘augmentation’ (the idea that human intelligence plus artificial intelligence needs to be more than the sum of the individual parts) actually occurs.

This isn’t the first time Davenport has treated a hot tech topic to make it digestible. Davenport’s milestone book, Competing on Analytics is now ten years old! When the book came out in 2006, data mining was still very much happening in dusty corners of the organization, sometimes leading to strategic insights to influence strategic decisions, but mostly becoming watered down when it came to influencing actual operational decisions. With a mission to bring analytics into the corporate mainstream and help make data science a boardroom concern, SAS even organized some very successful events with Tom here in Belgium.

The new book made me reflect on where things stand today. While analytics HAVE become an essential part of doing business, there is still a need to continually innovate as a competitive weapon. Today, there is more of everything (data, computing power, business questions, risks and most importantly, analytics consumers). The ability to scale your organizational analytical power is more essential than ever for staying ahead of your competition. And at the same time, in order to balance both the potential and risks of artificial intelligence, we all need to step up our game. Computers can and will do more intelligent work, but we also need long-term thinking about what and how we teach intelligent machines.

So with equal attention for what’s good for organizations today, as well as for planning the artificial intelligence augmentation of tomorrow, I wanted to bring attention to a few recent SAS whitepapers which address some key topics for meeting the growing demand for analytics.

• “Data Mining from A to Z: How to Discover Insights and Drive Better Opportunities” explains how the different flavors of predictive analytics work as intelligent filters for big data insights.
• “Redefine Your Analytics Journey with Self-Service Data Discovery and Interactive Predictive Analytics” discusses how approachable analytics enables citizen data scientists and business analysts.
• Finally, “Managing the Analytical Life Cycle for Decisions at Scale: How to Go From Data to Decisions as Quickly as Possible” outlines how organizations can generate “Decisions at Scale” to meet the growing appetite for analytics.

Reading these whitepapers will give concrete suggestions about making your analytics scalable to drive more efficient decision making. While reading them though, I suggest trying to also think about how these analytic process lifecycles need to be enriched with human insights.

• How do we develop models not only in line with organizational strategy, but with the organizational conscience?
• When establishing data labs for building predictive models, how do we think in terms of correct model behaviors, instead of correct model outcomes?
• When engaging machine learning, how do we ensure that such learning is holistic, not just maximizing expected utility from a logical perspective, but fully humanist?

Heady topics, but I’m convinced they are essential considerations if we as data scientists are to make positive contributions to artificial intelligence evolution. I’d be interested in hearing anyone’s views on this.