Machine Learning & Big Data Blog

Why does Gartner predict up to 85% of AI projects will “not deliver” for CIOs?

Chrissy Kidd
4 minute read
Chrissy Kidd
image_pdfimage_print

Earlier this year, industry research firm Gartner made an audacious prediction: that 85 percent of AI projects won’t deliver for CIOs. That means in 20 artificial intelligence projects, only three will succeed – 17 of them will fall short.

With all the buzz and talk about AI technologies, that’s a particularly surprising forecast.

Except maybe it isn’t. Let’s take a look at the reasons fueling this claim; Gartner surely has its reasons. We’ll also take a look at whether this outlook is any brighter in 2019.

2017 year-end predictions

Every year, Gartner polls thousands of CIOs around the world to get their takes on everything in IT: what’s working, what isn’t, which trends are busts, and which might have staying power.

Gartner’s word is powerful. A couple years back, the firm encouraged companies to focus on scalability as a way to ensure business growth. And this year, companies rose to that challenge, with a two-fold increase in companies reporting that scalability had been a top focus.

And so, in 2017’s fourth-quarter survey, Gartner asked about artificial intelligence. At the time, experts estimated that revenues for cognitive and AI systems could reach $12.5 billion that year – a nearly 60 percent increase over 2016’s revenue. But in that Q4 survey, only one in 25 CIOs reported using AI at all. That’s a mere four percent. So, what justifies the difference between the revenue prediction and the actual use of AI?

The answer lies in the nature of new technologies and our pace to adopt them. After all, though only four percent reported using AI, another 46 percent had plans to follow suit, whether in the short- or medium-term.

Early adoption for AI

These numbers indicate that we are certainly in the early adoption phase for artificial intelligence. While major tech giants like Google and Apple are researching and developing a whole slew of AI practices and technologies, the significant majority of companies, technology or otherwise, are lagging far behind.

Right now, it seems that most companies don’t know what AI can do for them. There’s lot of confusion around the technology – is machine learning the same as AI? What about natural language processing? Does embracing AI mean we’ll all lose our jobs sooner than later?

At this point, artificial intelligence seems like it’s more a matter of when than if. The adoption of such technologies is comprehensive and inevitable. Though AI itself can seem ethereal or just another buzzy trends, there are plenty of examples of AI already in action, from customer support and chatbots to recommendation engines and fraud detection to image recognition. It’s even possible for AI to help improve graduation rates or decrease repeat crimes.

Why will so many AI projects fail?

Risk and confusion are two main reasons for Gartner’s prediction about failure rates for AI projects. First, the decision to replace legacy systems and processes is not easy or quick: employees and customers alike know how your products work and what to expect from you. Considering how to invest time, training, and money in developing successful AI projects is a big risk – one that most companies can’t quite justify just yet.

Once companies opt for an AI project, confusion can wreak havoc. A general lack of understanding around all things AI means you may not have enough data, or that data may not be suitable to the project you’re considering. If your data isn’t good, your algorithms can’t be tested correctly – maybe you’re using the wrong algorithms for what you’re trying to solve. Any amount of understanding can fuel poor team management, but the more misunderstanding, the more likely the team is just wasting time.

Still, let’s say you have the problem and the data that may solve it. Finding the talent is tricky – because AI is not a smooth, single process or technology, there are currently few “AI experts”. The people who have such knowledge and skills, however, are in high demand and can command significant salaries. In a case like this, with good data but poor talent, will your company be able to trust what the AI is doing?

Outlook for AI in 2019

2019 could be a continuation of this early adoption phase for AI or a bellwether for change.

Early predictions indicate positive change. Gartner’s October 2018 survey of more than 3,000 CIOs shows that AI is now the most-mentioned technology, bumping data and analytics down to second place. Recognizing the way they asked the question changed from last year, now 37 percent of CIOs indicate that they are already using AI or that they have short-term plans to do so – a significant jump for one year.

Still, Andy Rowsell-Jones, a vice president and analyst at Gartner, preaches caution. CIOs may have an “irrational exuberance”, responding to the industry’s own push towards what AI technologies may offer us. Indeed, hardly a day goes by without a news article or press release touting AI at work. Perhaps we are embracing the idea that they should incorporate AI simply because they think they should.

Predictions for AI are expected to soar for a few more years, at least. Gartner forecasts that business derived from AI could reach $3.9 trillion in 2022.

Ready to get in on the AI action? A word to the wise: start with a core problem that’s essential to the entire company. Once you nail that down, you need to have a significant amount of data – so much that your employees aren’t able to analyze it on their own in any timely way.

In the coming years, AI won’t be the only path to business success; it’s simply one of them. AI is part of a third era of technology that encourages sound strategies, like scalability and security, with knowledgeable use of disruptive technologies like AI.

References

https://www.zdnet.com/article/close-to-half-of-cios-plan-to-deploy-artificial-intelligence-enterprise-solutions/#ftag=RSSbaffb68
https://www.informationweek.com/big-data/ai-machine-learning/ai-begins-to-infiltrate-the-enterprise/d/d-id/1330809

Get practical guidance for AIOps

IT operations teams have to work faster and smarter than ever to meet the demands of digital transformation. This ebook offers a practical, real-world look at ways artificial intelligence can improve the speed and efficiency of ITOM.
Download E-Book ›

These postings are my own and do not necessarily represent BMC's position, strategies, or opinion.

See an error or have a suggestion? Please let us know by emailing blogs@bmc.com.

About the author

Chrissy Kidd

Chrissy Kidd

Chrissy Kidd is a Denver-based writer who specializes in making sense of theories and new developments in technology, science, and medicine. Connect with her at http://www.chrissykidd.com.