Be with me for a while again and speculate about the current state of generative artificial intelligence (GenAI) apparently riding the new technology hype cycle defined by Gartner. The crazy fast pace of GenAI adoption we see since ChatGPT premiere is producing seemingly unlimited expectations about its future applications. The possibly disruptive capabilities of generative AI inspires scientists, media, business, and general public. Discourse about GenAI has become a trendy thing. Today AI is literally everywhere on the Internet and has already leaked into the regular small talks in the real world. It is a clear sign that, according to Gartner’s model of new technology hype cycle the phase, GenAI is currently the first phase of the cycle. The enthusiastic, awe-fueled popularity growth, which leads to the peak of inflated expectations. Most evidence suggests that we are not going to reach the peak soon, rather we’ll follow the bright, straight path up climbing high for a long time.

Everything is going up

Wherever we look all stuff related to GenAI explodes. Social media are spitting out news, memes, arguments, and trivia about artificial intelligence. The new events, symposia, congresses dedicated to AI generally ad to its applications in specific areas mushrooming abundantly. Trainings, courses, post-diploma studies are offered more and more aggressively. Linkedin – one social platform I am frequently on, is literally flooded with AI content. Users are sharing pictures, audios, videos, and posts created by recently launched GenAI tools. Many of users escalates in self-promotion as GenAI experts, tutors, evangelists or challengers. My feed on the platform looks like infected by GenAI FOMO (fear of missing out). Everybody looks like they are desperate to jump on the bandwagon of artificial intelligence. Outside social media the same trend makes corporation announce a new AI project every week and high schools and universities open AI-dedicated curricula. The technology itself is also on the rise. The capabilities of newly deployed models exceed the former models by far. The new model are made and launched frequently and new applications for them are found every day. Some of them immediately win users attention and approval. Chat GPT itself was the fastest-growing consumer application in history. The observation undoubtedly shows us that, if only Gartner’s cycle exists, Generative AI is in its first phase.

The challenge of the cycle phases

The Gartner’s cycle describes the process of new technology absorption by society. It means it is  sensitive to many factors changing the specific process and deviate it from the theoretical pattern. One of the best examples is the behavior of asset markets. The cycles of investor sentiment exist objectively. The consequence of bull and bear market phase applies to every asset, but yet nobody is extremely efficient in predicting the future performance of specific asset. It is just so hard to spot the moment when bulls give up and bears prevail (and vice versa). If only any investors knew that shares he bought are going to drop ... well but generally they didn’t. The same is with Gartner’s cycle. With a high level of certainty we know that one day our expectations toward GenAI become inflated and the technology bring us disappointment of some sort. The other big things will arise and big players, investors, and media will turn their interest to them, leaving Gen AI on the slide to the trough of disillusionment. But how long we will wait for this moment. Are we at the start or at the very end of the growing phase of the hype cycle?

In my opinion, there are warning signs already here

Despite the impression, that everything is moving up in the Gen AI I see some signs of disappointment. It is subjective view, of course, but probably those kind of personal, biased, human-specific opinions are today the last stand of where humans can exercise their expertise better than algorithms. Where do I see those warning signs? Firstly, Nvidia shares are not growing like crazy anymore. They have been in a clear horizontal trend since May 2024. Some can argue, that Nvidia is already in the next phase of market cycle, some say it is only a correction of the otherwise impressive growth. This is a point - we can’t tell if Nvidia will rise or slide in next month and whether its price reached only a local bump or historical maximum three months ago, in June 2023. We will know it, however, when time passes. The same is with GenAI. I hear more frequent skeptical voices about practical application of the technology in business. I meet many doubts about European AI Act and experts saying it is going to slow down the adoption of GenAI. During professional events it is easy to find a debate, fire chat or keynote sharing the view that GenAi is overhyped. All those signs could mean that we are somewhere close to the peak of the cycle. On the other hand there is still a lot of positive energy and actual progress in the area. Maybe those warning signs means something, but maybe not. I my opinion  everybody who confidently says that GenAI is already on the slide to disappointment just grasps at straws. I do think, however, that,in my opinion, the question of whether GenAI is on its ride up or starting to slide down is valid and meaningful today.

Gartner’s hype cycle enhanced by the question in subject. Source: Wikipedia.org