Future of Decisions: Replace Gut Instinct With Artificial Intelligence

Future of Decisions: Replace Gut Instinct With Artificial Intelligence

An effective digital transformation means minimizing the guess work, and automating decisions through AI.

Robert Heller once said one “should never ignore a gut feeling but also should never believe it’s enough. In the digital age, the late British management journalist’s words are taking on even more meaning.

By some estimates, the average adult makes about 35,000 conscious decisions every day or nearly 13 million per year. Some are basic, like what should we eat today (that’s the most frequent one)? Others are more complex, like should we hire or lay off staff? Should we enter or leave a market? Is it time to make a big strategic move or not?

In most cases, every decision considers two things: relevant and available data and gut feeling. It’s a fairly reliable formula that has stood the test of time and inspired an infinite number of quaint sayings extolling the value of trusting one’s inner voice. But with data and digitization enabling businesses to move faster and more competitively than ever before, human beings are finding it difficult to keep pace.

That’s why automated technologies, such as artificial intelligence (AI), must play a more central role in helping people make quicker, actionable, accurate, and accountable decisions in the digital world.

“We need help,” says Aera Technology CMO Ram Krishnan. “Things are moving faster than we can possibly handle. There is just too much data to process, and we can no longer afford the mistakes that occur when relying on gut instinct. We are now at the point where we need AI to rescue us.”

The Limits of Human Decision Making

The trouble with traditional methods isn’t that they don’t work very well in the digital age. There is so much online data these days that those companies able to quickly assess and act upon it are generally able to leapfrog competitors that cannot. That means every business leader must now make decisions faster than they normally would.  But the required pace of decision making is beginning to eclipse the limits of our human capability.

The human brain, you see, is capable of processing about 11 million bits of information per second. Most of that action happens without us even realizing it. We take in our visual surroundings, hear a sound and quickly forget it, or notice a fragrance as it wafts by. Our conscious minds, by comparison, can only handle about 40 to 50 bits of information per second. That simply won’t cut it in todays’ fast-paced business environments where a recent Gartner survey found 65% of decisions are more complex and involve far more stakeholders and choices than just two years ago.

Unfortunately, to circumvent that complexity, many of us take cognitive shortcuts and rely on subconscious data (our guts) rather than facts and figures. It’s a dangerous and error-riddled path.

For example, imagine being an overworked HR lead. You’ve always seen young, white, male, hotshots in executive positions and haven’t had much experience with older women of color. When resume pictures from this latter group land on your desk, unconscious biases might creep in that lead you to look at other candidates and toss worthy resumes in the trash – a discriminatory practice that could lead to lawsuits.

Things are moving faster than we can possibly handle. There is just too much data to process, and we can no longer afford the mistakes that occur when relying on gut instinct. We are now at the point where we need AI to rescue us.

AI Over Instinct

Today, with the speed of business being what it is, companies can ill afford such mistakes. And the only way to effectively get on top of them is by minimizing gut instinct and automating decisions through AI.

What does that entail? Essentially, it means you deploy technology to collect, aggregate, analyze, and recommend the best path to success in the shortest amount of time. Human beings still make the ultimate decisions. But the digital machine does the grunt work to guide them to the best outcome.

This capability is important at a macro-level for organizations because some of the most common and costly mistakes occur when one business unit’s decisions disagree with another’s  Consider, for instance, a salesperson plugging an overly optimistic annual sales forecast into a company’s customer relationship management (CRM) system. That forecast gets rolled up into a monthly forecast that goes into a planner that ultimately leads to a certain amount of merchandise being produced and stored for eventual shipment. When sales aren’t completed and product begins piling up on warehouse shelves, the manufacturing and procurement side of the house takes note. If it happens again, their guts might tell them to disregard the salesperson’s forecasts and produce less product. Such snap decisions could render them unable to meet demand because of low inventory, ultimately resulting in lost revenues if the forecast happened to be more accurate this time around.

The Need for Decisioning Audits

Having an AI system in the mix, in this case, would have more accurately tracked actual sales, communicated proper production needs, and allowed those manufacturing and procurement leads to make more data-driven decisions related to the sales forecasts. What’s more, when that eager-beaver salesperson inflated the numbers, the system could’ve warned that person they could be going too far based on history.

“When decision makers are confronted with information that does not seem to align with their gut feelings, they often have strong reactions,” says Raul Castanon-Martinez, a senior research analyst for 451 Research. “This illustrates the necessity for quantitative and AI-based decision making. It also points out the need to enable systems of record for tracking decisions and how well they’re made.”

AI systems are particularly adept at tracking decision-making histories, from ideation to result.  If fed the right guidelines  – goals, strategies, milestones, costs, expected benefits, timelines and so on – they can objectively and cost-effectively track and analyze projects in ways human beings simply cannot.

“If you asked human beings to do this, you’d have to decide who would be tracking such decisions. And then you’d always have to worry about whether they have the time to diligently do it - and if they can keep their own biases and motivations out of the mix,” says Krishnan. “What you really want is to know is how well the team executed against the decision and whether it yielded the right results from a goals and objectives standpoint so you can do better next time.”

Establishing AI Guardrails

That’s not to say people can hand off everything to machines and forget about it. Indeed, AI and machine learning (ML) platforms are only as good as the data they are fed. Even with the best data, machines can jump to the wrong conclusions. They also struggle with out-of-the box or creative thought.

Humans, therefore, will still need to partner with them on some tasks. What does that mean? It means that after initial setup, we can hand over menial and redundant tasks to them. They can even make certain basic decisions, like responding a common customer inquiry or running a set of programmed responses for getting lights back on during a power outage. But for more complex issues, machines will most likely be limited to crunching data and making actionable recommendations.

Even with those guardrails, humans will still need ways of understanding how AIs arrive at  conclusions. We cannot, after all, see how computers process data. At the same time, we’ve been conditioned over the years to fear Terminator-like machines wiping us out or robots eliminating our jobs. To fully trust our machine partners, we will need what the U.S. Department of Defense (DoD) calls “Explainable AI.” Still conceptual, the idea is that as AI systems evolve, you build-in capabilities enabling them to fully and  transparently explain their decisions or actions.

Early Stages – For Now

Even with such guardrails in place, being able to fully trust machines to make solid recommendations or trustworthy decisions will take time. AI is still in its infancy, and the technology is far from perfect.

But Krishnan believes most obstacles will be overcome sooner than later because the value proposition for AI is so high and the technology is advancing so rapidly. Gartner, for example says half of enterprises will have AI orchestration platforms by 2025, up from fewer than 10% in 2020.

“AI will advance because it must,” he says. “The world keeps saying, ‘decide faster, faster, faster.” And if you can’t attribute and measure the impact of your decisions well enough, you’re doing your business and customers a disservice. AI will empower organizations to fine-tune decision velocity for the future.

From Process Automation to Decision Automation: How to Make The Next Step in 2022
Taylor Wills
From Process Automation to Decision Automation: How to Make The Next Step in 2022
Are Touchless Supply Chains Already Real?
Taylor Wills
Are Touchless Supply Chains Already Real?
The Five Imperatives Of The Fourth Industrial Revolution
Pascal Bornet
The Five Imperatives Of The Fourth Industrial Revolution