It should be obvious to anyone in the technology field that artificial intelligence is being hyped to unrealistic expectations at the moment. But this makes it no different from any other technology that society has generated through the ages â€” from the cloud to virtualization to the service oriented architecture (remember that?), and all the way back to the personal computer. There are probably old newspaper clippings touting the ability of steam engines to create flying machines and rockets to the moon.
While much of hype is wishful thinking, sometimes there is a bit of truth to it. Past technologies have certainly reshaped the world, although not necessarily in ways that early boosters had envisioned.
Starting the cycle
AI is at the nascent state of this process, which Gartner has encapsulated in its Hype Cycle. The latest release has technologies like AI-augmented software and generative AI rapidly closing in on the â€œPeak of Inflated Expectations,â€ while more esoteric applications like quantum machine learning and AI-driven innovation are just starting out at the â€œInnovation Triggerâ€ stage. So far, none are even close to the end-stage of the cycle, the â€œPlateau of Productivity,â€ let alone the dreaded â€œTrough of Disillusionment.â€
Perhaps this is one of the reasons why most IT executives are still looking at AI with jaundiced eyes. A recent survey by KPMG showed that a good 75% of top decision-makers view AI as more hype than reality at the moment; half say that AI is developing too fast for them to keep up. A key detriment to AI is that the implementation hurdles are extremely high, which is what we should expect from something that is designed to remake enterprise data environments on a granular level.
At the same time, there is a distinct lack of understanding of what AI actually is, how it works, and what it can ultimately do. This is one of the consequences of excessive hype; in the case of AI, expectations range from full automation of all data functions to clear and unfettered insight into highly complex and seemingly intractable data-driven processes.
If the past is prologue, however, these and other misunderstandings will clear up once AI is operational and people learn how to use it. The goal for enterprise executives at this point should be to shift the focus away from the hype and more toward the value AI can bring to the business model.
According to Alex Ubot, vice president of Solution Engineering at Dataiku, most organizations encounter three key roadblocks when implementing AI: governance of internal assets, management of talent, and difficulty of designing a clear technological roadmap. To overcome them, organizations will have to do the hard work of identifying the business use cases for each form of AI under consideration (and there are quite a number of them already), then implementing an inter-departmental working environment designed to align both the objectives and the way forward in order to shift away from the hype and focus on the value that AI brings to the business model.
Part of the team
This leads to the most salient point about AIâ€™s transformative impact on the enterprise: It is not a standalone, plug-and-play technology. In fact, it requires the enterprise to transform to a certain extent before it can have an appreciable effect on digital processes.
Nacho De Marco, CEO of tech solutions firm BairesDev, pointed out on Forbes some of the key steps organizations must take in order to derive real value from AI. Failure to define the proper users and use cases for each AI implementation tops the list, followed by neglect of the quality of data used to train AI models. In other words, you canâ€™t just throw AI into the enterprise and hope for the best. It must be carefully and strategically implemented, and then just as carefully and strategically monitored to ensure it is delivering on its promises.
The essence of technology hype is the practice of overpromising and underdelivering. AI does offer a lot of promise, but the implementation challenges are substantial. In the final analysis, organizations should treat AI the way it should have treated new technologies all along: Focus on finding the right solutions to real problems, not on acquiring the newest, shiniest tech in the channel.