Why Human Decisions Must Increasingly Move to Machines

Artificial intelligence isn’t new. But how companies are using it today is. As such, it requires every leadership team to take a close look because AI is going to separate the companies that dominate from the ones that fall behind over the rest of this decade.

To be sure, AI has been a key element of our digital ecosystem for a better part of the last 60 years. It has evolved significantly from systems that inform (from the 1960s to the early 2000s), to systems that advise and alert (last decade’s gold standard), to systems that act on our behalf.

AI that does our work for us is today’s Holy Grail. AI-infused systems that halt a misfunctioning manufacturing line or step on our car brakes in a panic stop have become staples of company and personal lives. However, the advent of generative AI large language models like ChatGPT has raised the curtain on AI, democratizing access to its ever-expanding potential. As a result, business leaders are commanding their teams to infuse all types of AI and machine learning into products and services.

This week we published a study of more than 300 U.S. executives examining this phenomenon. We looked at how their companies in three U.S. industries (transportation, healthcare and financial services) were infusing AI into their business. We found that companies are using AI today not just to give users information they weren’t aware of, and advice on decisions they must make. Increasingly, companies are using AI to make and execute decisions that humans can’t make quickly enough.

Our research uncovered early examples of companies that are gaining big benefits from AI-infused systems. One of them is Progressive Insurance. The huge property and casualty firm last year brought a new feature to its Snapshot smartphone telematics app. The feature, called Accident Response, uses the digital sensors in smartphones to determine if a customer has been in a major accident and (if so) to alert local emergency medical services and towing. The feature can also file a claim within 10 minutes after confirming an accident.

Or consider UPS’s DefenseDelivery AI-infused service. It helps shippers determine the probability of a package being poached from a recipient’s porch, saving all from the heartbreak of a stolen delivery.

But your company doesn’t have to be the size of UPS or Progressive Insurance to use AI in these ways. Much smaller companies that we surveyed were also generating outsized benefits. We studied what the companies with the greatest benefits from today’s AI did differently than those with the smallest returns.

The first lesson is that company leaders must think about not if, but when and where they should infuse AI into their products and services. Four other practices are crucial to consider:

• Design your AI-infused systems to make decisions with great consequences and that must be made in extremely short time windows. Look at your highest impact products and services and create intuitive systems that follow modern design principles. In fact, use AI itself to embed AI in your products and services (e.g., coding assistants, requirements gathering bots, automated testing tools, etc.). But keep humans in the loop to monitor and if, necessary, countermand machine decisions.
• Realize the more you design the system to take control, the more data, higher-quality data, and more current data you will need. Look closely at your data hygiene practices across all data types: structured, unstructured and semi-structured. Make sure the data is accurate enough to train your large language models and enable your company to take real-time actions on them.
• The more you rely on the system to make and execute key decisions, the more your C-suite must be involved in system design. Top management must fully understand the benefits – and especially the risks — of shifting tasks from people to machines. Overlook this and be prepared for organizational blowback.
• Focus on cycle-time reduction more than on cost reduction. Our research revealed that the best companies at using AI-infused apps focus more on reducing the time it takes to execute consequential decisions. It could be which sales leads to pursue, and not pursue. Or helping service reps answer complex customer questions. Or they could be helping your strategic planners get faster at deciphering new market needs and competitive threats.

As AI keeps evolving, you must rethink your market and operational strategies. You will be faced with ethical issues (e.g., potential bias and responsible use) and how much work to give people and the machines that you’ve assigned to them.

From our research, we are ready to discuss how your organization could think about embedding AI into its products, services and business processes. As we found in our research, I predict you will see a rich set of opportunities to grow the top and bottom lines faster.

To learn more, please read our research report “When and Where the Machine Should Rule ”.

About the Author

As Sparq’s Chief Executive Officer, Monty Hamilton leads the executive team and drives the overall strategy. Sparq is a unique digital engineering company focused on helping its clients realize their digital transformation strategies and creating thousands of technology careers in places where they otherwise would not exist. Monty has helped grow the firm through two private equity investments to become the largest privately held digital engineering firm with colleagues across the US and Latin America. He is a sought-after speaker on the outsourcing and domestic sourcing topic and has recently been featured on CNBC, BBC, NPR radio and at various industry conferences including IAOP, Gartner, Digital Georgia and others. In addition, recent articles depicting Sparq’s innovative outsourcing model have appeared in Business Week, CNN Money magazine, CFO magazine and CIO magazine.