Don’t Be Shy with AI: It’s Time for Cultural Change

Despite the current AI excitement, I expect many companies to make limited progress in the year ahead. Hoping for the hype to pass and the implementations to become easier, many leaders will keep their head down and focus on projects they feel more comfortable with. If I am right, then this increases the rewards available to anyone ready to fully embrace AI’s capabilities.

Generative AI is still poorly understood, which is unsurprising given the pace at which it is moving. But slow adoption creates a real risk of falling behind and a few AI experiments won’t be enough to make an impact. To see its full potential in the enterprise AI needs to be rolled out in many different business processes. For this to happen, leaders must focus on significant cultural change.

Why AI Demands Cultural Change

Like any technology, AI alone cannot deliver meaningful transformation. Leaders know that real change will only occur if the environment supports it. The problem is that today there is just as much fear and uncertainty around AI as there is excitement.

It would be naïve to suggest that none of that fear is well-founded. There are security measures that must be addressed, and the simple fact is that jobs will be lost to AI. But other jobs will grow, and AI will help people to get more done and do less mundane work. Erik Brynjolfsson looked at an AI-powered call centre and found out that average productivity improved by 14% very quickly, and productivity went up by 34% for the newest and the least-skilled workers.

AI is Creating New Jobs, Not Just Taking Them

In the industrial revolution new technologies created jobs as well as replacing them. There are signs that history may be repeating itself and some research suggests that AI will in fact create more jobs than it eliminates.

Nicholas Thompson, CEO of The Atlantic recently shared a view on this. He tells that for a long-time translation was one of the canonical examples of jobs that might not exist as AI became more powerful. The theory was that it’s a simple enough task that eventually machines will be able to do what humans did before. Nigel shares several examples of why this theory may not be correct.

The Atlantic itself provides one of those examples. “We used to not employ anybody to do translation, it didn’t work with our economic model.” They then had the idea to translate some of their work into Spanish, use AI to do the first pass and use someone to check and make sure it was up to standard. “The result is that we now have more translated stories than we did, and more work translating than we had before.”

So What’s Left for Us? The Maker and the Checker

Just as in the Atlantic example, humans retain an important role. AI is the maker, and humans are the checkers. A New York Lawyer famously used ChatGPT to perform legal research and identify cases, but his court filing was later found to reference example legal cases that did not exist. If he had simply checked the AI output before using it, he could have quickly caught the mistake.

It would be easy to think that the lesson from that story is that AI cannot be trusted, but the lesson is only that AI must be leveraged correctly and in a way that avoids hallucinations. Including a step to check AI output does not remove its benefits. I often think of AI as an exoskeleton for the mind. The human is in control but is empowered by AI.

Embracing AI Means Encouraging Creative Destruction

“Creative destruction” is a concept in economics identified by Joseph Schumpeter in 1942, saying that change is inescapable and that old ways of doing things are constantly being disrupted by new innovations. If this mindset is encouraged, organizations can create a positive cultural shift towards AI adoption, fuelling productivity gains. Leaders will benefit greatly from defining their desired culture around AI, communicating it out to the business and creating initiatives to deliver on that goal.