Exclusives

Artificial intelligence (AI) experts from companies including SHI and Advanced Micro Devices (AMD) helped separate the hype around generative AI from the reality about it on Oct. 12, during the webinar “Beyond the Hype: A Practical Guide to Assessing, Piloting, and Scaling Generative AI.”

The panelists were hoping that viewers would come away from the webinar with “some practical knowledge [about] generative AI’s capabilities and how you can realistically, practically apply that to your business to extract real value,” John Moran, senior product manager-automation and applied AI at SHI, said at the start of the webinar.

“Obviously, there’s a lot of information out there, and we want to kind of demystify some of that,” he noted.

At SHI, “my responsibilities are mainly internal facing, unlocking the power of generative AI at our business,” he said. “But there are not a lot of well-lit paths and blueprints for how to do this. So, I’ve naturally … focused on helping customers really travel the same path that we have and educate them on the pitfalls and challenges that they’ll likely face and that we faced so you can overcome them and start actually extracting value from this technology.”

Right now, with generative AI, “obviously we are in a massive hype cycle – probably the largest one of the last 20 years,” Moran said. “There is an incredible amount of enthusiasm and excitement but there is also an equal amount of confusion [and] misguided, unrealistic expectations that could potentially lead to wasted effort. And SHI and the folks on the [webinar] today have seen firsthand the challenges and the pitfalls that exist. And they are very real. There is a lack of well-lit paths. There is not a clear path to value yet for … generative AI, or at least it’s not readily available.”

There are also “security risks – some known, some unknown, [and] just misinformed, unaware expectations about what this can actually do and how simple it is,” according to Moran. “The majority of these problems are really spurred from just a lack of understanding and a lack of experience. Not only is the path undefined but the value that is actually addressable by … generative AI is really exclusive to five core business functions: Those being sales, marketing, software engineering, customer operations, and product R&D,” he said.

Moran pointed to a study by McKinsey that basically said, ‘yes, there is value, but it’s exclusive to a few functions’ – and we’ve actually seen that internally as well.”

Therefore, he said: “Not only is the path undefined. It’s also very narrow. That being said, the amount of value potential that’s available from generative AI and what we’ve seen is unavoidable…. Generative AI specifically expands the total impact potential of AI beyond just on its own, beyond any other form of traditional AI” or machine learning.

Meanwhile AMD’s strategy around AI is “based on three [strategic] pillars,” according to Shiva Gurumurthy, senior marketing manager at AMD. “The first pillar is leadership products. That’s our entire product portfolio and how it maps to the problem of … AI in general.”

He went on to say: “The second pillar is really the leadership architectures that we bring to the table and which is also a key link between the hardware and the software ecosystem that’s out there.” The second pillar also includes IP efficient use of silicon.

And, he added, the “third pillar is our outreach to the ecosystem, to what we call the pervasive AI software.”