The “Agentic AI” Hype Cycle Is Insane – Don’t Normalize It

Two weeks ago, I published an article, “Agentic AI Is The New Vaporware.” I reasoned that “agentic AI” is a hype term that repackages pie in the sky AI ambitions, but does not allude to any particular advancement that might achieve them. The buzzword amplifies the overpromising narrative that we’re rapidly headed toward a great leap in autonomy. Most uses of “agentic” do not refer to any novel technical methodology and the ambition of increasing autonomy is not new – even as the word falsely implies otherwise on both accounts.

The article performed fairly well. Only four of my 26 Forbes articles have received more views.

But I doubt the article made a great dent – I don’t expect the hype to quiet down overnight. The Gartner Hype Cycle currently has “AI agents” sitting at the “Peak of Inflated Expectations.”

Unfortunately, tech seems to wear the peak hype position like a badge of honor – despite the connotations that the next move will be a grave downfall into the “Trough of Disillusionment.” The cycle is seen as a right of passage. There is a belief that hype, even when overzealous and misleading, helps propel technology.

The Gartner Hype Cycle Actually Promotes Hype

The Gartner Hype Cycle famously depicts a presumed trajectory for each new technology, from inception to maturity. At first, expectations rise as a new technology gains traction. But the hype typically goes too far, reaching the “Peak of Inflated Expectations.”

With agentic AI, we’re right at that precipice, ready to tumble down to the “Trough of Disillusionment.” Agenic AI’s plunge could be worse than most. The disconnect between hype and reality seems to only be increasing, with so much more attention on a colorful story than concrete, deployed value.

Overselling is costly; when the executives catch on, there’ll be hell to pay. If we continue on the current course, the fall from grace will be even steeper and the disillusionment even deeper than usual. It could even be central to a third AI Winter, an era of diminished excitement and funding.

But when Gartner positions a technology at peak hype, it worsens the hype rather than tempering it. The firm’s Hype Cycle normalizes – even glamorizes – the irrational, costly overpromising and overselling that hype entails. The message is that these ups and downs are not only inevitable, they’re helpful for accelerating the development and adoption of a technology.

I disagree. There’s a detrimental complacency in passively accepting, “That’s just how things work in the tech industry.” There’s complicity in agreeing to an arbitrary amount of exaggeration and overpromising – or, as Gartner calls it, “Inflated Expectations” – as copacetic. Sure, the message of hype’s unavoidability is a bit somber, but Gartner’s Hype Cycle also indulges in the excitement, effectively celebrating the exuberant trendiness. This amounts to a celebration of poor expectation management.

It’s true that excitement helps get people moving. But must we promote fiction in order to generate excitement and productivity? I’m too much of a humanist and technology optimistic to think yes.

AI Hype Takes The Cake

This problem of hype begeting hype is most intense for AI. After all, AI hype is the worst hype. It propels a narrative that, relatively soon, the tech will be capable of taking over human labor wholesale. It’s a false promise of extraordinary levels of machine autonomy, fueled by an enticing, fictional story of escalating “intelligence” that begins to propel itself like wildfire to transcend the entire range of human capabilities.

But Gartner rewards each trend for… being a trend. For each technology it includes on its up and down ride, the Gartner Hype Cycle serves as a one-size-fits-all stamp of approval. It’s incapable of issuing a more dire warning, even if warranted – you never see its “rollercoaster” curves depicted as more or less extreme, nor its “Plateau of Productivity” particularly low or even hugging the floor. But for some technologies, you certainly should.

The “agentic AI” hype is even more egregious than most AI hype. The words themselves – “agent” and “agentic” – magnify the grandiosity. Crediting machines with “agency” doubles down on AI’s core mythology and original sin, the anthropomorphization of machines. This sells an implausible, unsupported narrative that we’ll soon see unprecedented, significant new levels of computer autonomy.

The “Agentic AI” Hype Wave Is Only A Continuation The “GenAI” Hype Wave

Gartner has bolstered the “agentic AI” hype wave even further by announcing it as its own technology with its own wave. Although “agentic AI” has assumed the guise of being a unique technology unto itself, the hype term does not actually refer to any particular methodology or advancement. Rather, “agentic AI” is a continuation of the broader genAI hype cycle that we’ve been in the midst of for almost three years. Gartner appears to have not recognized this. Their latest report suggests that genAI has already slid halfway down toward the “Trough of Disillusionment,” separate from “agentic AI,” which is positioned atop the hype peak.

I’m afraid that the current, singular wave of “genAI/agentic” hype – spawned from and based on large language models and other forms of generative AI – has far from peaked. Pivoting to its newer terminology helps rejuvenate the wave; “agentic” has simply picked up genAI’s mantle. After all, the term “agentic AI” has come to refer almost exclusively to uses of LLMs.

Since it normalizes overpromising, one could hope that the Hype Cycle itself will someday become its own victim. With a bit of luck, it will dip down to its own “Trough of Disillusionment” and stay there.