It has only been around a few years, but the question is already emerging: Has the hype around generative AI reached its peak?
As businesses look at gen AI today and understand its impressive innovation potential, it’s also a heavyweight system that comes with significant costs and complexities, according to David Linthicum, research analyst for theCUBE, SiliconANGLE Media’s livestreaming studio.
“We’re … trying to figure out what it is and what we can use it for,” Linthicum said in a recent AI Insights and Innovation series podcast from theCUBE, which features the latest news, trends and insights in artificial intelligence, with a focus on generative AI. In this edition, Linthicum looks beyond the hype around generative AI adoption to examine emerging options for businesses.
Hype around generative AI: The good, the bad, the ugly
The conversation around gen AI has shifted from just focusing on its potential. Now, it’s about its real-world challenges and limitations
“Business leaders are increasingly concerned with issues like practicality, costs and the challenge of integrating it into existing business processes,” Linthicum said. “With the rise of ChatGPT people can see the value in it. “Obviously, it has had a lot of value it can bring to applications, to business … because it’s able to generate new uses of information.”
However, as Linthicum points out, the very features that make generative AI so powerful also contribute to its limitations.
“It takes a huge amount of data, processing power and money to train these systems,” he said. “So, as business leaders look at the applications or the use cases for generative AI, they’re often unclear about the use cases for their particular organization.”
Generative AI is not going away, but there may be better options for many businesses, Linthicum added.
“We understand the power of gen AI, and the value of it is getting clearer with all the new releases of LLMs out there,” he noted. “But businesses, CIOs, CEOs, boards of directors are…trying to figure out if this is…actually going to provide business value for them.”
There’s a lot of hype around generative AI, along with inaccurate or misleading information from AI-generated news articles, or summaries that contain factual errors or lack of context. Normally, gen AI technologies “go a mile wide and an inch deep … because they only have information that they’ve been trained to provide,” according to Linthicum.
But when it comes down to day-to-day operations, many CIOs are finding that traditional technologies might serve their needs just as well, if not better, at a fraction of the cost. There’s a lot of interest in gen AI, even among employees, but clear use cases are still hard to identify.
“They also see when they look at the cost … they just don’t have enough budget to deploy some of these huge generative AI systems … for their particular business case,” Linthicum said.
Is agentic AI the next big thing?
Many business implementing gen AI systems are not necessarily looking at other options, such as agentic AI, which is based on a less-expensive technology leveraging flexible AI agents that can be embedded with certain business processes, not necessarily entire systems. Linthicum sees a shift in the industry toward lightweight deployment of these AI-based systems that can leverage smaller, more specialized AI models for specific tasks.
“Agentic AI represents a return to a more tactical use of AI, where the focus is on solving specific business problems with smaller, more manageable systems,” Linthicum said. “These systems don’t require the massive resources of LLMs, making them more accessible to a broader range of businesses.”
“Those have been around for a long time,” he noted. “They’re going to be cheaper and therefore deliver more value to the business.”
Linthicum suggests that businesses should explore how agentic AI and other lightweight technologies can complement or even replace generative AI in certain scenarios. What the marketplace is seeing is the normalization of the hype around generative AI, finding real applications for it and how to make it financially viable for companies.
There are other challenges around “governance, strategic integration, ongoing innovation … and with security. All these sorts of things are problems … [that] still need to be solved,” Linthicum added. “Generative AI is not going away — this is exciting technology…and it’s going to have some value. It’s just being oversold right now.”
Businesses need to be realistic about gen AI’s limitations. The goal should be to use it in ways that make sense for the business, not just to follow the latest trend, Linthicum advised.
“Right now … businesses are aligning themselves in other directions in some instances. They’re looking at other options, agentic AI, non-AI systems, workflow-based systems [and] robotics process automation (RPA) … to provide more tactical uses of AI. And I think that’s the smart thing to do,” Linthicum concluded.
Be sure to check out more of SiliconANGLE’s and theCUBE’s coverage of agentic AI in these articles: Gen AI is passe. Enter the age of agentic AIand Wait, will gen AI really pay off? Inquiring investors want to know
Here’s theCUBE’s complete episode of the “AI Insights and Innovation” podcast: