Understanding the Hype Cycle
Now the Gartner Hype Cycle ™ — as you know, which is right here, and as I say, we will give you a copy to look at — is a longstanding way that we study how trends evolve. There’s a trigger that happens and people become interested in this technology. It reaches a Peak of Inflated Expectations. Eventually, it comes off of that peak into what’s called the Trough of Disillusionment, where the work happens, and then eventually it might actually reach the Plateau of Productivity.
Generative AI in the Trough
The thing about the AI Hype Cycle is how fast things are moving through it, partially because of the amount of investment that’s been happening over the last year and a half to two years. Now, GenAI, as you’re probably feeling, generative AI is past the Peak of Inflated Expectations and starting to go into the trough. In all of my conversations with clients around the world, this appears to be really true. You’ve heard me say before that the Trough of Disillusionment is not like this dark, dangerous place, but it is the place where we figure out how to make something work or not, what it is or not. It’s where the hard work takes place, where the other dependencies are, and so on. Generative AI clearly is going into that place.
Emerging Trends and Plateaus
But the acceleration of generative AI is pulling a bunch of other things along with it. There are things that are still working their way toward hype, things like multiagent systems. It’s quite early on the curve. One of the things that we're learning right now is that one of the ways to integrate across models — say models from different major providers — is that integration is going to be accomplished by agents talking to one another. There was an announcement recently from Microsoft and SAP about the fact that Microsoft Copilot and Jewel, which is SAP’s agent, can now talk to one another to interoperate across the two different environments. We’re going to see this more and more and more.
You take a trend like multiagent computing, or agentic computing, which is another way of thinking about it, and it is actually going to accelerate pretty quickly because the need is there. I also want to point out that there are a number of things that have reached that Plateau of Productivity already. Things like computer vision — there's a long discipline of mature work there. Computer vision is key to multimodal generative AI, but it itself, as a part of AI, is very mature and already productive. Other things like knowledge graphs are hitting that Plateau of Productivity, and so on.
There was an interesting question when we posted this on LinkedIn a little while ago about the Hype Cycle itself. Somebody asked, “Well, don’t things ever go away off the Hype Cycle?” Yes, they do. Sometimes when something goes up through the trigger and across the Peak of Inflated Expectations and hits the trough, it never comes out. It does happen. We make that assessment regularly to say, did this technology go away? Was it obsolete? Or did it fold into something else? I expect some of that to happen with the AI Hype Cycle as we go along.
The AI Hype Cycle tracks the evolution of AI technologies from initial interest to widespread adoption. It highlights how generative AI has moved past the peak of inflated expectations and is now entering the trough of disillusionment, where real-world applications and challenges are addressed. The video also discusses mature AI technologies like computer vision, while identifying emerging trends like multiagent systems and agentic computing.
Learn more: Building a Value-Driving AI Strategy
