When a new iteration of technology breaks through, we often see the same pattern repeat: Big ideas, enthusiasm, investments, and initial mass adoption. But then?… [insert silence here]… the hype fades, and focus shifts. Will the story be the same for generative AI? We don’t think so… 

Not long ago, VR, AR, and the Metaverse attracted significant amounts of venture capital—this has decreased by 77% since the peak in 2021. Observing that about 20% of all global venture capital during the first half of 2023 went to companies working on AI makes it clear that the hype is in full swing. 

Gartner also places generative AI at what they call the “Peak of Inflated Expectations” level. However, all new technologies go through this phase, and either they sink to the bottom (only to resurface in a new form a few years later) or they reach a plateau where they are valuable, enduring, and can be implemented in reality. 

However, this cycle with generative AI feels a bit different due to its potentially disruptive impact on broad areas of “knowledge work.” It’s not just about “edge cases” for individual industries or functions but a technology that offers an enormous opportunity space for support, optimization, and quality improvement through generative AI, relevant to many functions across almost all industries. 

If we turn our gaze back to Gartner’s Hype Cycle, they estimate that generative AI is about 2 years away from reaching what they call the “Plateau of Productivity.” 

However, we believe that we should look at these two years as dog years… but the opposite. Dog years are counted as seven “human years,” but with the progress made in generative AI over the past 12 months and the amount of investment flowing into AI, we should consider “AI years” as half a “human year,” maybe even less. Moreover, we believe the difference lies in the accessibility of generative AI. We no longer need big servers in the basement; we may not even need to code a single line to implement AI in many of our workflows. We use “Software as a Service” solutions via HCI or API, often paying only for what we use, and therefore, we think we should already “jump on the train” and start embracing this technology.