The AI Industry has a ‘Jack and the Beanstalk’ problem

TL;DR: If the process and data don’t change, the results usually won’t either, no matter how advanced the technology might be.

Most of us learned Jack and the Beanstalk as a success story.

A family is struggling. The one asset they rely on stops delivering what they need. Jack is sent to sell it and come back with something that will help. Instead, he trades it for something speculative. The magic beans grow, and everything works out.

That’s usually how the story gets told. But if you look at it from a business perspective, it reads a little differently.

Jack didn’t have a clear plan. He made a decision under pressure, with limited information, and traded something real for something that might work. Things got dicey, and he got very lucky.

Sometimes, that type of gamble pays off. Most of the time, it doesn’t.

That dynamic shows up regularly in conversations I have with clients. Over the last decade, there’s been no shortage of “magic beans.” Mobile apps. Blockchain. The metaverse. Digital transformation initiatives. And now, AI. Each wave of disruption brings with it similar expectations: faster execution, lower costs, competitive advantage, access to capabilities that weren’t available before.

If you step back, the pattern is consistent: a new technology shows up, the upside gets framed in broad terms, and the pressure to adopt builds quickly.

In a lot of cases, the urgency behind those decisions isn’t coming from an immediate business constraint. It’s coming from a mix of fear of falling behind, pressure from the market, and a sense that something needs to change after a series of stalled or unsuccessful internal efforts.

Forget “Fee, fie, foe, fum!” and try “FOMO, FUD, and AAUGH!” on for size.

I regularly speak with leaders at companies that have been trying for years to improve their workflows or modernize systems. In many cases, these projects took too long, didn’t deliver what was expected, or never really got traction internally.

It’s no surprise that leaders have seen AI as a shortcut to bypassing some of that work. The assumption is that this technology is different enough that it can skip the effort required to fix the underlying issues. AI might be a step change in capability, but without changes to process and data, most organizations will see results that look a lot like what they saw with earlier waves and promises.

Many of the assessments we’re doing for clients reveal that technology itself isn’t the issue. Instead, organizations don’t realize value because:

• The business problem isn’t clearly defined.
• Different teams don’t agree on how the process is supposed to work.
• The data that feeds that process is inconsistent.
• And once something new is implemented, it’s not always clear who owns the result.

When those conditions exist, adding a new layer of technology doesn’t change the outcome. It just changes how quickly you get there.

From the outside, change often looks like progress, but inside the business the outcomes usually aren’t changing in a meaningful way. That’s usually when companies start adding more tools, hoping something will stick.

In my experience, the companies getting real value from these investments are approaching the problem differently. They start with the outcome they’re trying to improve. They understand how that outcome is produced today. They look at where the process breaks down and where effort or time is being lost. Then they decide where automation or augmentation fits.

This isn’t a new approach. It’s the same kind of operational discipline companies have always needed. It just tends to get skipped when there’s urgency around adopting something new.

The appeal of the Jack and the Beanstalk story is that it turns a complicated situation into a single decision. Trade the cow, then climb the beanstalk. Everything works out from there.

That’s just not how businesses actually operate in practice.

Outcomes are produced by a combination of people, process, and data. Technology interacts with and supports that system. It doesn’t replace the need for it to be well understood.

Before committing to a new technology, you can improve your chances of success if you explore these questions first:

• What specific business problem are we trying to solve?
• What outcomes would indicate success?
• How is that outcome produced today?
• Where does the current process break down or create inefficiency?
• Is the process consistent enough “as is” to support automation?
• What data does process depend on and how reliable is it?
• Who is accountable for the result once a better solution is in place?

If those answers aren’t clear, introducing something new tends to add complexity without improving results.

That doesn’t mean companies should move slowly. There are situations where moving quickly makes sense. What usually separates effective decisions from expensive activity is whether that groundwork has been done.

There’s always going to be some level of uncertainty in these decisions. That’s part of running a business. The difference is whether you’re making a decision with a clear understanding of what needs to happen for it to produce the desired outcome, or whether you’re relying on hope that it will work itself out.

The story of Jack and the Beanstalk focuses on the ups and downs (literally) that led to a lucky result. In practice, it’s usually more useful to look closely at the decisions that precede and predict outcomes. Before trading something real for something speculative, it helps to be clear about what problem you’re trying to solve and how the new approach is expected to create value.

And if it turns out that magic beans (e.g., AI) are the solution after all? At least you’ll be better prepared for any giants (or golden-egg laying fowl) you happen to meet along the way.

This article was originally published on Mark Richman’s LinkedIn – follow him there for more insights for leaders.