Cracking the code: Why more companies are focusing on AI projects

Karen Roby interviews a Gartner analyst about why some companies are doubling down on artificial intelligence and machine learning projects.

Cracking the code: Why more companies are focusing on AI projects

A new study reveals over half of organizations have some AI or ML project in the works, and three-quarters of those companies plan on doubling the number of projects that they'll be deploying within the next year. TechRepublic's Karen Roby talked with Jim Hare, an analyst with Gartner, about the study. The following is an edited transcript of their conversation. 

Jim Hare: I think some of the key takeaways that we found out of the survey was that over half of the qualifying organizations have some type of AI or ML project currently deployed, and three-quarters of those have plans to roll out more projects in the next year, two years, and three years. It looks like they're planning almost doubling the number of projects over the coming years, which to us we found pretty profound. Of course, these are organizations that are more on the front line, the leading edge of using AI and ML. They're the ones that are getting the value and have cracked the code, as I'll call it.

The other things that I think are key takeaways was, what's the primary reason that they're wanting to use AI or ML? There are two things that we came out of that. The first one was automated task, and the second one was using AI and ML to improve customer experience. This is very consistent with what we've seen before where really customer experience, customer engagement, understanding who that customer is, delivering a much more personalized experience is one of the top reasons.

The second is with the automated task is to try to gain more efficiency, both in terms of improving business processes as well as improving the decision-making by the employees of an organization. So, those are a couple of the key takeaways we found that I found, actually, fairly interesting.

Karen Roby: How broad of a sample did this take into consideration in terms of the size of the companies involved and also the industries?

Jim Hare: This was around 100 respondents, and they're members of what we call Gartner Research Circle, which is essentially a set of Gartner clients that have agreed to participate in different surveys. And then, what we did is narrow the scope to say these had to be individuals who responded who are familiar with an organization and its AI or ML initiatives or may be involved in the process of choosing particular technologies or specific use cases.

So that said, of these hundred, the mix was actually across the board. If we look at it from where they were geographically located, almost half were out of EMEA, 46%, 36%, or about a third, out of North America, and the remainder were in APAC and Latin America.

And then, when we look at it from a company size point of view, about 20% were in that 1,000 to 2,000 employee range, and the remainder, or the bulk of them, were in the 2,000 to 10,000, but there were some organizations that responded even less than 100 employees, and then some organizations even had 50,000 employees or more. 

From an industry perspective, the two primary industries for these organizations were the government space and manufacturing, and that was 20% for each of those each, and then the services and banking side with like 15 and 14% respectively. And then, we did have respondents from insurance, retail, and so on. So, to me, it was a fairly broad representation, but the bulk of them coming much more from the government and manufacturing side.

SEE: How smart tech is transforming the transportation industry (TechRepublic Premium)

Karen Roby: Did you guys expect to see a jump this big in terms of the number of companies that are looking to add even more AI projects in the near future?

Jim Hare: We did not, so that was one of the things we found fairly profound. Those that cracked the code, those that have actually started deploying AI and maybe have four or five projects, have figured out now how to scale it and use it in their organizations. So, going from, say, four this year all the way to 35, 36 in the next three to four years, we found pretty amazing.

Karen Roby: This may be a broad question here, but what is it about AI and ML that is so compelling, or why is it proving to be so beneficial that companies want to jump on board and put more of these projects into play?

Jim Hare: I think the misnomer is people hear the term AI and they think it's going to get rid of people automatically. In other words, it's going to replace what humans are already doing, and hence, the reason for the term artificial intelligence. It should be called augment intelligence. Most of the value from AI is going to come from using the power of the machines to work alongside individuals to do the things that are tedious and very time-consuming, things that can be automated, things that require a lot of number crunching, but then deliver those insights or address some of those mundane tasks to free up the humans to use a lot more of the human power. So, the augmentation of using AI and ML is really where a lot more of the value is going to be delivered.

Karen Roby: As with anything, the more tailored a program, the more expensive it is, so we're talking about a really good chunk of change here that's being spent on this technology.

Jim Hare: The more you build a custom solution, one, you have to find the skills, people, to really build and even deploy and manage that on your behalf, so it can be very time-consuming when finding those resources and building and deploying those models. I think people should think about AI as really a spectrum of capabilities that can be packaged in different ways. There'll be some things that you build that are custom, perhaps because, again, you can't find an out-of-the-box solution, or you feel it's something you need to own as a differentiator for your organization.

But for a lot of other users and use cases inside the organization, there's nothing wrong with buying a prepackaged application that already has AI capabilities that can be used to solve a specific discrete problem. And then, there's even packaged AI capabilities increasingly being added to enterprise applications, such as your ERP, or supply chain, or CRM applications. Those are AI capabilities baked directly into the business processes. And it's not that one's better than the other. Organizations need to be thinking about a portfolio of AI capabilities that they'll use inside their enterprises.

Also see