Whether you are exploring the world of AI for the first time or you are deep into integrating it, the technology has been evolving rapidly and making its way across industries in new and exciting ways. It can be extremely helpful to have feedback or insights from those that have already walked the walk with integrating AI into their products or ecosystems.
During our yearly virtual customer summit, Rustici: The Gathering, we invited colleagues and friends that have delved deeply into AI to hear their take on where to start, how to get team buy-in and much more. Our panel of experts included Stephen Kalnoske, Lead Product Development Engineer of SANS Institute, Gary Lamach, Director of Learning Consulting for AICPA & CIMA, Gavin Beddow, VP of Product for PeopleFluent and Chris Tompkins VP of Business Development at Rustici Software to guide us through their insights. We’ve attached a preview below along with some of the most insightful takeaways from their session, and we hope you find them just as helpful.
What problems were you trying to solve by using AI?
Stephen Kalnoske of SANS Institute
One of the big problems we were trying to solve was understanding what was in our content. What would happen if we had a client who came in and wanted to take cybersecurity courses on a specific topic, there was no centralized way to suggest content. Now, with AI, we can break down all our content to know what’s inside, removing the previous two-week-long process of control mapping through lab manuals and other content.
Gary Lamach of AICPA & CIMA
We’re close to a thousand courses, so the content curation piece and categorization are definitely something that AI helped us with. Even tagging our content, going through and mapping to fields of study has been a tremendous help to us. Everything was manual beforehand, using our course descriptions to identify content that was in there. Now, we can focus on those creative endeavors and make more interactive content.
Gavin Beddow of PeopleFluent
We’re looking at how we can solve real human problems. Specific areas of focus for me are looking at how we can reduce some of the items that are more labor-intensive tasks and leverage greater insight into the data products. This involves reducing common things that people do in our products day in and day out and using AI to augment and accelerate that content.
How are you approaching AI with making effective content?
Stephen Kalnoske of SANS Institute
Firstly, we basically never let it create anything in a void. The other thing is to ensure that the material is accurate. This involves being very transparent organizationally about where we use AI and where we don’t. For anything where AI is generating learning objects, we’re tagging them as AI-generated. When we feed them into other systems where people are looking at them, it’s very clear what AI is and what is not AI, so we can keep that in mind as we work with all this data.
Gary Lamach of AICPA & CIMA
We don’t do pure content creation with AI. It’s really a tool to assist the human, you know, the proverbial human in the loop. We still rely heavily on subject matter experts to review, evaluate and rewrite the output of AI. I liken it to Clippy in Microsoft Word. It’s a tool that helped us write, but it didn’t necessarily do the work for us. We’re just at a point where those tools can do more for us now and assist us even more.
Gavin Beddow of PeopleFluent
We do need that human intervention. We can take the collective knowledge that an organization might have and pass it on to existing content courses. However, building an eye on the industry or learning knowledge and not just the general knowledge that an adult may have is more of a difficult task. We should avoid mass content creation and rely wholly on AI because it lacks that human creativity moment, too, and it certainly can lead to content mediocrity.
If there’s a lesson you wish you had known before you began tinkering with AI, what is it?
Stephen Kalnoske of SANS Institute
I would suggest making sure everybody uses all your AI tools in a good way from the beginning. This comes with plotting out what we train people on with using the tools and how we approach AI in the workplace.
Gary Lamach of AICPA & CIMA
My piece of advice that I can give everyone is to never get locked into the current state of whatever tool you’re working in. Make sure that there’s always room to adapt as the technology evolves and ensure that your processes and your protocols internally allow for that flexibility. You never know what tomorrow’s going to bring with AI.
Gavin Beddow of PeopleFluent
I would suggest not being too wedded to specific technologies. The technology bedrock that we’re using today will probably have to be iterated in the future, as the technology is just moving so quickly. New capabilities are getting unlocked all the time, which we will absolutely want to exploit, and we have to evolve our fundamentals behind the scenes to give us the best chance in the future.
We really appreciate our colleagues taking the time to explore the concepts of AI within eLearning and we hope you enjoyed their answers. If you are interested in learning more and getting the full answers, check out the recording where they go in-depth to these questions and more. Plus, sign up for our email newsletter for news on our work with AI plus other AI-related eLearning content.
