Lockdown Lessons for the Intelligent Enterprise

Published on INFORMS Analytics Magazine (Joseph Byrum)

Augmented intelligence can provide the flexibility that is key to success.

Adversity is a great teacher, which means we ought to be able to learn quite a bit from the COVID-19 pandemic. Smart executives are already hunting for the opportunities opened by the societal changes forged during the hard times of 2020. In the near future, the intelligent enterprise will use artificial intelligence (AI) to become more adaptable to situations so that they’re ready to not just survive the downturns but also make the most of the new opportunities.

As one example of how things are different, it’s certain that we’re never going back to the drudgery of morning commutes and set-in-stone 9-to-5 business hours. Valuable employees have seen they can conduct a business meeting over Zoom and no longer need to hop onto those red-eye flights to distant cities, wasting both personal time and company resources for an in-person meeting that might only last 30 minutes. Knowing this, employees who recognize their worth will consider flexible work policies when calculating the quality-of-life impact of prospective job offers. As they’ve seen a more attractive alternative, the old ways will not survive – at least not at the companies that want the best talent.

That’s one of the small ways in which companies will have to adjust in the years ahead, and changing course is critical to survival. The ability of a large enterprise to adapt itself to the changing marketplace determines its ability to succeed.

Failure to Adapt Can be Fatal

The history of business is filled with examples of thriving enterprises that missed the signs that the environment had fundamentally changed and were reduced to insignificance over time. Sears tumbled from its position as the nation’s number-one retailer through the 1980s [1] only to find itself in bankruptcy court in 2018. The department store giant’s rise began in 1888 with Richard Sears selling watches and jewelry through the mail [2]. Yet, less than a year before Jeff Bezos came up with the idea of selling books through the mail from orders placed on the Internet, Sears ditched the catalog that made it famous [3]. Sears’ profit soared to record levels in the short term [4] while the opportunity to dominate online sales by applying a century of mail-order experience to the then-new World Wide Web was lost. Throughout the 1990s, Sears also failed to innovate in terms of logistics, allowing rival Walmart to seize the top spot. Today, Sears hangs on as the nation’s 75th largest retailer [5].

It’s hard to blame Sears. Nobody can predict the future, and an upstart like the Amazon of the early ’90s had more freedom to experiment with new business approaches than a firm that, at the time, managed $54 billion in annual revenue and 350,000 employees [6]. When you’re already on top, it’s easy to see the danger of messing with the recipe for success.

Enter AI

What AI can do now, that it could not do two decades ago, is provide a framework to assist executives at businesses of any size in making those tough judgment calls about when it’s time to do something different. AI provides the intelligent enterprise with the situational awareness it needs to recognize the need to adapt to fast-changing circumstances.

A machine’s ability to ingest and process information is unrivaled, but data collection alone is not particularly valuable to a busy executive. The great leaps we’ve seen in AI development have come on the software side as natural language processing and data crunching algorithms are beginning to make the information collected actually useful. Expert systems are being built to incorporate the in-house expertise from the organization’s best minds so that the AI platform can evaluate and prioritize information that affects the organization. Systems like this can run simulations on potential outcomes and determine not just what is happening, but also attempt to formulate answers to the question of why it’s happening.

This is what an executive needs before making a decision about, say, selling off one line of business and investing in another. This is what distinguishes augmented intelligence systems from the forms of simple machine learning that tend to enjoy the media spotlight. The primary purpose of augmented intelligence is to act as a decision support tool. It informs the human decision-makers who, based on the information received, make the choice about what to do.

How to Design an Augmented Intelligence System

An effective approach to building a decision support tool is to follow the lead of military strategists who developed a mental framework that boosts a pilot’s ability to make fast, but accurate, choices in the heat of combat. These pilots found through experience that whoever acts first has a massive advantage, and they used what they called the OODA Loop to react faster than their opponent. The same techniques can help businesses gain an advantage over their competition in the safer, but no less fast-paced, marketplace environment.

The first step in the mental framework is to “observe” (the ‘O’ in OODA), which means to take in all the data unfiltered, resisting the urge to apply your interpretation to what it might mean. If preconceived notions interfere with this process, important signals can be missed – especially ones that are often confusing or incomplete.

Once the data are gathered, the next step is to “orient” your thoughts to the information, formulating theories that would explain the unfiltered data. There would be multiple theories, spawning multiple potential actions. The “decide” stage is to choose which theory best fits the facts, and the “act” stage is where the plan is implemented. The whole process loops back to the beginning, with an unfiltered evaluation of the situation, so that the effect of the choice can be evaluated.

If a business looks at the pandemic environment, sees data suggesting a change is needed and implements that change, the most important thing is to confirm whether the change is producing the desired result. If not, it’s time to adapt and try something else. The mental framework encourages constant evaluation and reevaluation of changing situations, which is critical to success. The role of AI is to collect the data, model the results of potential courses of action, and present recommendations to the executive who will make and execute any decision.

Ensuring the executive’s awareness of the current situation is evaluated around the clock is a job perfectly suited for AI, and the intelligent enterprise relies on augmented intelligence to be ready to pivot as needed to survive and thrive.

Many companies, particularly large ones, are accustomed to “sticking with what works” long after the opportunity for a meaningful change has passed. AI in itself is not a solution to that problem; it’s one part of the solution. Companies need a culture that embraces change for this process to pay off, and for some this becomes the greatest challenge to overcome.

It’s equally important to have the right talent on hand to evaluate and act upon the information provided by augmented intelligence systems. The good news is that work continues on both fronts as companies continue to make great progress toward creating the intelligent enterprise of the future that will learn and make the most of both good times and bad.

References

1. https://www.sun-sentinel.com/news/fl-xpm-1991-02-15-9101080852-story.html 
2. http://www.searsarchives.com/catalogs/history.htm 
3. https://www.baltimoresun.com/news/bs-xpm-1993-01-26-1993026112-story.html 
4. https://www.chicagotribune.com/news/ct-xpm-1994-02-09-9402090302-story.html 
5. https://nrf.com/resources/top-retailers/top-100-retailers/top-100-retailers-2020-list 
6. https://archive.fortune.com/magazines/fortune/fortune500_archive/full/1995/