Crawl with Analytics Before Running with Artificial Intelligence
As we head into 2017, artificial intelligence (AI) is an emerging technology trend that is generating increasing buzz among business leaders of both start-ups and well-established companies. From virtual assistants to image recognition to self-driving cars, we’re only beginning to scratch the surface of how AI-related technology will impact our daily lives and transform how businesses operate. IBM’s Chief Innovation Officer Bernie Meyerson optimistically predicts 2017 will be “the year of the solution as opposed to the year of the experiment.”
The worlds of artificial intelligence and Internet of Things (IoT) are poised to collide. This year’s CES conference in Las Vegas was packed with all kinds of smart devices touting new AI capabilities. Artificial intelligence will transform everyday objects just like electricity did more than a century ago. As Wired Editor Kevin Kelly stated, “Everything that we formerly electrified we will now cognitize.” Just as electricity pulses through all businesses today, AI-related technology will also begin to reshape how businesses work. Box CEO Aaron Levie foresees AI technology introducing “a dramatic change in how enterprise software is designed and how enterprise software takes advantage of all the data that’s in our platforms to produce way better outcomes for customers.”
While the tech giants such as Amazon, Google, Apple, IBM, and Microsoft have been investing in AI technology for years, the question is whether 2017 is the year when we begin to see more mainstream adoption. While the marketing hype will be everywhere and we’ll see advancements in 2017, the technology will not reach a tipping point this year. A recent Forrester survey found that while 58% of companies were researching AI, only 12% were currently using AI systems. While we won’t see the widespread adoption of artificial intelligence technology for a few years, that doesn’t mean you shouldn’t start preparing your organization for an AI-powered future today. In many cases, it might be just the type of competitive advantage an industry disruptor needs to get a leg up on slower moving competition.
Recently, some companies have suggested skipping traditional analytics and jumping right into AI. Last October, Babson professor and analytics expert Tom Davenport reported how he encountered several companies at Salesforce’s Dreamforce conference that wanted “to ‘leapfrog’ over traditional business intelligence and analytics capabilities and go directly into AI-related environments.” He noted how one manager admitted her nonprofit organization had used almost no analytics in the past but was excited to jump into AI. Another technology executive mentioned to him, “We’ve had enough bar charts. Nobody has time to digest them anyway. We want our analytics to tell people what to do.” With many organizations still struggling to embrace data, analytics is still an area that all companies need to address before diving into artificial intelligence. Bypassing analytics is not a shortcut to AI because analytics maturity is a key milestone on the path to being successful with AI.
As your company advances its analytics maturity and capabilities, you’ll see increasing AI augmentation as you progress up through the analytics levels. (image/Brent Dykes)
There are many different levels to analytics, starting with descriptive analytics (reports, dashboards) and diagnostic analytics (drill-downs, ad-hoc queries), which focus on understanding the past. The next three levels turn their attention to the future and are increasingly augmented by AI technology. While the potential of predictive, prescriptive or cognitive/autonomous analytics is exciting, there’s some essential groundwork that’s needed at the lower levels. If your company hasn’t mastered the basics of analytics yet, it’s fallacious to believe you skip ahead with AI and without a solid analytics foundation.
Without the transparency that analytics provides, it will be difficult to judge the results of any artificial intelligence system. We’re already beginning to see examples of poor decisions being made by algorithms and data models with little insight into their rationale. Analytics holds AI accountable and can help optimize the effectiveness of AI systems over time. In addition, many of the problems that can derail your success with analytics will also sink your AI efforts. Therefore, it’s important that your organization learns how to crawl and walk with analytics before running with artificial intelligence. Here are some areas where analytics helps to clear a path for AI success:
- AI is powered by the same fuel as analytics—data. If you’re not collecting the right data or lack faith in the quality of your data, it won’t be magically corrected with AI. Just as your analytics systems rely on accurate, complete data so will any AI technology. Increasingly, artificial intelligence may be used to improve data quality, but it won’t compensate for data that is fundamentally corrupt or unreliable.
- AI is also dependent on many of the same business processes that impact analytics. For example, data privacy is a sensitive subject with consumers, especially as companies collect more and more data on their customers. Without sound data privacy best practices that are a byproduct of a mature analytics program, AI technology could inadvertently misuse customer data in ways that could damage goodwill and brand perceptions. Analytics essentially highlights the upstream and downstream processes that could also be impacted by AI.
- Artificial intelligence will also depend on having skilled people with both domain and data expertise. Their insights into the business, key processes, and data are essential to implementing AI technology effectively and they will be responsible for maintaining and improving AI systems over time.
- AI relies on having a data-driven culture in place so that it can be fully embraced. If your managers and employees frequently question or ignore insights from your analytics tools when making decisions, AI won’t alter this resistance. However, if they become accustomed to relying on and trusting your analytics data, they will welcome opportunities to lean on artificial intelligence to further augment their analytical capabilities.
- AI will face many of the same organizational roadblocks that can impede success with analytics. If certain managers and teams already feel threatened by the greater transparency created by analytics, good luck trying to introduce anything like AI that further threatens their perceived power and influence. Breaking down internal political barriers with analytics can clear an easier path for future AI adoption.
Before you pursue AI technology in 2017, evaluate how far your analytics capabilities have progressed to date. Just like crawling is a key developmental milestone for infants, analytics is equally important to aspiring data-driven organizations that look to embrace artificial intelligence. A lack of analytics or data maturity will delay your organization’s ability to benefit from emerging AI technologies. With artificial intelligence potentially ushering in a new industrial revolution, it’s imperative that each organization learns to crawl and walk with analytics in earnest before seeking to run with AI. The data-driven companies that do so will be able to quickly outpace their rivals that failed to adequately prepare for the coming AI revolution.
This article was written by Brent Dykes from Forbes and was legally licensed through the NewsCred publisher network.