Panel explores the good, the bad, the ugly, and the unknown of AI
Five local leaders offer their insights during an event hosted by the Knoxville Technology Council.
Edmon Begoli, our former PYA colleague, said that he was breaking out in a sweat just thinking about all of the bad ways that artificial intelligence (AI) could be used.
Those comments came during a panel discussion last Thursday afternoon hosted by the Knoxville Technology Council (KTech). Begoli, AI Systems R&D Section Head at Oak Ridge National Laboratory, joined four other leaders from area businesses who discussed the good, the bad, the ugly, and the unknown of AI. They were, in alphabetical order:
- Greg Elin, Senior Principal Engineer and Evangelist for OSCAL at RegScale;
- Darrell Jenkins, Chief Information Security Officer at Clayton Homes;
- Gary Jackson, Director of Artificial Intelligence and Blockchain at CGI; and
- David Resseguie, Director at the PwC Labs Innovation Hub.
During a roughly 70-minute session that KTech Co-Founder John McNeely moderated, the five discussed everything from their concerns about the positive and negative aspects of AI to the ways that their enterprises are responding to and preparing for the impacts that will occur.
All seemed to agree that the technology was in its infancy and much was still to be learned although, in response to an intriguing question from an audience member, Begoli said the opportunity to have an individual’s own AI model was coming “faster than you think. All the components are there. The next few months are going to generate so much innovation.”
Less than two days later, Axios AM reported that Rep. Ritchie Torres (D-N.Y.) plans to introduce legislation that would require the products of Generative AI to be accompanied by a disclaimer. The article in Saturday’s edition noted that the “bill is the latest in a wave of new legislative efforts to regulate AI as Congress grapples with the emerging technology’s massive potential — both for societal advancement and harm.”
One of the many insights that the panelists offered was one from Jackson – an interesting analogy for how he was approaching AI at CGI.
“We want to use it like interns,” he said, explaining that individuals should not freely embrace the technology any more than they would let interns operate without any supervision. “Check in like you would with an intern.”
In the case of PwC, Resseguie noted that the firm’s clients are asking, “How do we transform ourselves using AI?”
Elin said that RegScale has developed a Generative AI tool that the company is offering to help write better code for compliance, the major focus of the fast-growing entity that is headquartered in the Washington, DC area but has its R&D operations in Knoxville. With so much of the innovation at RegScale occurring locally, we took note of one of his observations in particular.
“This is a tremendous opportunity for East Tennessee to lead in answering these questions (about AI),” Elin said near the end of the panel discussion. Those questions cover the spectrum – from how enterprises elect to incorporate AI into their daily operations to policies they implement, new training approaches for their employees, ways they are better protecting their data from AI-enabled hackers, and how to best use AI for good.
The latter point was emphasized by Resseguie who said, “Let’s make people know how to use it properly.” Later, he added, “It’s going to fundamentally change the way we train people.” How so? Resseguie explained that, from a positive perspective, AI will allow enterprises to automate low-level tasks, freeing up time for employees to focus on higher-value work. However, that will also undermine the reality that lower-level work allows newer employees to learn from their mistakes. Determining how that will change training programs is something that clearly must be considered.
Jenkins said that Clayton Homes is not “inherently blocking AI.” Instead, the Blount County-based business has a team of human resources and legal professionals developing policies that will guide future training.
“We need to understand the limitations of AI,” he said.
On the point of training, Elin said, “If people don’t know how to use technology, they’re going to use it in bad ways,” emphasizing the clear linkage between the organization’s culture and its policies related to AI.
What’s the greatest risk? Begoli said he believed topping the list was misuse to fool humans and attacking systems, resulting in data breaches that are not only very costly from a financial perspective but also to an enterprise’s reputation.