Scientifically, the popular model is already on shaky ground for AI experts. Hilgenstock points to studies that show that the hype cycle lacks prognostic validity. While technologies such as smartphones or cloud computing have simply grown steadily over a decade without ever passing through a “valley of tears”, other trends such as virtual reality or blockchain develop in unpredictable, recurring waves.
AI is a wave movement
In his opinion, the one-dimensional model fails, especially when it comes to artificial intelligence. The history of AI is not a single hill, but a series of independent waves. The “AI winters” of the 70s and 90s were followed by ever new bursts of innovation: from expert systems to statistical machine learning to today’s generative models (GenAI).
What is special about the current situation is the simultaneity. While the world marvels at ChatGPT, other forms of AI such as forecast models or recommendation systems have long been part of everyday, productive life in companies. And the next wave is already building on the horizon: “Physical AI”, in which intelligence moves directly into robots, vehicles and machines. For Hilgenstock, these different levels of maturity cannot be squeezed into a single curve.
