Getty Images/iStockphoto

Deriving value from generative AI with the right use case

The technology will be valuable to tech vendors. For users, a return on investment will depend on the applications as well as whether enterprises choose to build or buy their models.

BOSTON -- While fears about the potential of generative AI to replace many human jobs continue to rise, technology vendors and customers are still seeking the correct use cases for the new technology.

According to IDC, spending in AI technology is expected to reach $521 billion by 2027. That is spending on AI software and infrastructure by tech vendors and enterprises.

However, tech vendors and customers are just starting to figure out the real value of the technology and how investing in generative AI can lead to positive ROI.

The value comes not only from monetary gains, but also from productivity gains, IDC analyst Philip Carter said during a presentation at the IDC Directions conference in Boston on March 14.

Finding the right use case

"It's a double-sided coin in the end," Carter said. "The more you deliver that experience for your customers, the higher the chance that you pick up a chunk of that $521 billion. But it starts with the use case."

The use case helps determine the ROI and productivity level. Generative AI can also help workers by relieving them of the time they spend performing many mundane tasks, Carter said.

Determining the right use case is what GE Digital is currently considering, according to senior product management executive Prasad Pai. The software company provides software and services to organizations in the manufacturing, utilities and power generation sectors.

"We can position our solutions [and] we can start to think about where do we integrate GenAI into our offerings," Pai said in an interview with TechTarget Editorial.

IDC analyst Philip Carter presenting at IDC Directions.
At IDC Directions, analyst Philip Carter discusses how monetizing AI is about finding the right use case and deciding the buy-or-build strategy.

The opportunity within data centers

For Vernon Turner, an adviser for sustainability and IoT at tech consulting firm Cognizant, generative AI holds great opportunity for IT data centers.

For example, he said, when there is a problem in an IT data center, instead of using a service menu that directs users to help, generative AI steps in.

"The data or information that will be generated can be used to make operations more efficient," Turner said in an interview with TechTarget Editorial. That is hard to do today because of the lack of machine intelligence within data centers, he added.

We're turning IT from a call center into a service center. IT has an opportunity to be a value center.
Vernon TurnerAdviser, Cognizant

Generative AI can also help move to a greener infrastructure in data centers, Turner continued.

"GenAI will be a part of the solution by giving operations a chance to know where energy is being consumed," he said.

Corporate sustainability officers tend to think of data centers as a small piece of the infrastructure and overlook it, but with generative AI, CIOs and chief sustainability officers will need to work together to apply the technology correctly.

"There's so much information that's sitting there," Turner said. "We're turning IT from a call center into a service center. IT has an opportunity to be a value center."

Building and buying

Other than finding the right use case, there's also a shift in the question of buying or building generative AI tools.

"There is an architectural conversation that needs to take place, which is around the build and buy [question]," Carter said.

While enterprises have long faced the buy-build dilemma with tech tools, with generative AI, that is no longer so as much, Carter added. There's also a middle ground.

"Depending on the use case and its strategic competitive advantage, as well as your availability of in-house resources, you could be training, or you could be building your model or doing that with a partner," he said. He noted that most organizations take the approach of fine-tuning existing models, especially if they're industry-specific.

For Pai, taking advantage of the possible opportunity available in the generative AI market will mean that many tech vendors will have to invest in building some models, especially domain-specific ones.

"It's very clear that GenAI is going to be an ecosystem solve for the industry," he said. "It can't be one company doing it all. So, there has to be a balance that we strike between how much do we build versus how much we partner, and even buy, for that matter."

However, in finding the use case for generative AI, the worry persists that technology will replace many workers.

While the fear is understandable, the technology has yet to develop to that point, said Shital Shah, IT director at specialty glass company Corning, in an interview.

"Big companies take a long time to get to that point where they can automate everything," Shah said.

She added that part of the process of finding the value of generative AI is having good-quality data.

"If you don't have good-quality data, none of this will matter," she said.

Esther Ajao is a TechTarget Editorial news writer and podcast host covering artificial intelligence software and systems.

Dig Deeper on Enterprise applications of AI

Business Analytics
Data Management