radical Insights.

Weekly Research and Commentary on the Future of Business and Technology.

Decoding the Hype: Unmasking the Misconceptions of the Gartner Hype Cycle.

Jun 11, 2024

The other day, a client asked the presenters in a multiday program covering diverse technologies (in the old Singularity University days, we would have called them “exponential technologies” – nowadays I find the term rather misplaced, but that’s a post for another day) to rank each technology on the famed Gartner Hype Cycle.

For those unfamiliar, the Gartner Hype Cycle is “a graphical representation developed by the research and advisory firm Gartner to depict the maturity, adoption, and social application of specific technologies. It is a tool that helps organizations understand how emerging technologies will evolve over time and their potential impacts on businesses.” (To stay with the theme of hype, I asked ChatGPT for the definition.) It looks like this.

It’s one of those nifty tools that compresses a complex subject into a simple infographic. It makes people feel informed and gives them a way to discuss emerging technologies. However, it is also hugely flawed (this is not a criticism of the methodology or the tool itself but due to the way it is often presented plus a specific aspect of its graphical representation). Allow me to elaborate…

For starters, the x-axis (time) doesn’t have a scale. Worse, by combining different technologies on the same graph (which is often how people use the Gartner Hype Cycle), it suggests that all technologies move on the same timeline and scale. However, the time it takes for a specific technology to move from the “peak of inflated expectations” to the “trough of disillusionment” varies greatly. Yet they both appear with the same time scale on the graph. To be clear, this is not a flaw in the Hype Cycle itself (as Gartner deliberately doesn’t put a scale on the x-axis) but a common misinterpretation, aided by the visual design of the chart (since the x-axis usually has scales in most other graphs).

Check out Gartner’s 2023 “Hype Cycle for Artificial Intelligence”.

“Prompt Engineering” and “Artificial General Intelligence” appear close together (both nearing the “peak of inflated expectations”). However, I believe we can all agree that prompt engineering is becoming mainstream (and already has to a certain extent) far quicker than AGI (which might or might not happen at all).

Secondly, and this is an even bigger problem, the way the Hype Cycle is drawn suggests that every technology will eventually move through the five phases and come out victorious. This is far from the truth. Some technologies will just peter out and never climb up the “slope of enlightenment.” I remember when NoSQL databases were all the rage and considered the end of good old SQL. After more than a decade, we found some specific use cases for NoSQL, but interest is waning, SQL keeps growing, and NoSQL’s growth has become negative. Other technologies will quickly be superseded by their successors and never reach the “slope of enlightenment.” I still have some MiniDiscs somewhere, which were rendered obsolete by MP3 and later streaming services. Lastly, many technologies never make it at all. My drawer is full of widgets that were the “next big thing” and flopped. Every company that has been around long enough has a graveyard of projects that never recovered from the dark recesses of the “trough of disillusionment.”

This is to say that one should be careful with simple, two-dimensional views of a complex world. Not every technology will succeed, no two technologies progress at the same speed, and the future remains (deliciously) uncertain.