According to Gartner, we’ve passed the peak of inflated expectations for AI, and we’re heading towards the “trough of disillusionment” — organizations are struggling to get business value out of AI even as companies continue to invest billions in new models.
I believe one of the causes of confusion is that there are two very different AI trends going on at the same time. There’s the hare, using AI for personal productivity. And the tortoise, using AI for corporate processes.
First, the hare: Depending on the survey, around 30% of professionals say they use AI at work on a regular basis. And all the data says that the more people use AI, the more valuable they find it.
But time saved by individuals is not necessarily the same as ROI for the organization– there are lots of anecdotal reports of people using AI to do their jobs faster, but not telling their bosses, because they don’t want to get told off for using it with corporate data, or they are worried that their bosses will give them extra work, or even replace them with an algorithm.
And that brings us to the tortoise. The vast majority of AI value will come from embedding it directly into your existing workflows and systems, in ways that it “just works”, in the same way we’ve gotten used to algorithms in our regular lives, from autocorrect to Google translate, or automated Spotify playlists.
But it’s going to be harder. We’re moving from the idea that “AI will change everything tomorrow” to “It will change everything eventually, but only if we do it right.”
All the analysts agree that we’re now entering into the tortoise phase of AI, moving from exploration to construction, and that 2025 will be the year of Enterprise AI adoption.
According to Gartner, 92% of CIOs believe AI will be implemented in their organizations next year—more than any other technology. Unfortunately, today, only 20% of companies are approaching AI strategically.
And so, unsurprisingly, Gartner says that having an AI strategy is the #1 priority for CIOs in the coming year, closely followed by data and analytics. And when trying to implement AI, the number one obstacle is legacy, non-cloud systems, followed by a lack of skills and data complexity.
You all know the fable: the tortoise might be slow, but it wins the race. And in this case, the race isn’t just about who adopts AI first—it’s about who gets it right.
So will your organization be ready to cross the finish line?