Why Traditional Training Can't Keep Technical Teams Current - with Nicola Lyons of Andela

Daniel Faggella (0:12)

Welcome, everyone, to the Emerge AI in Business Podcast. Today's guest is Nicola Lyons, Senior Director of Talent Operations at Andela. Nicola discusses why traditional event-based training can't keep up with the technical skills that now shift every six to 12 months, and why enterprises need continuous assessment-based learning instead of one-time courses. She walks through how personalized learning paths target the specific gaps individual engineers have, whether that's inefficient AI tool use or hesitation to adopt new tools at all, so teams can build real capability instead of sitting through generic content built for everyone. Today's episode is sponsored by Andela. According to Edison Research, 79% of Americans age 12 and older listen to online audio monthly. That is an estimated 228 million people. For busy executive audiences, podcasts offer a rare opportunity to capture 20 or more minutes of attention with VP and higher-ranking leaders in America's largest enterprises. Emerj reaches millions of those listeners every year. To learn more about how recipes drive pipeline for other AI brands, download our media kit at emerj.com/av1. That's emerj.com/ad1.

Now the conversation with Nicola.

Nicola, welcome to the show.

Nicola Lyons (1:52)

Hi, thanks for having me, Yolande.

Daniel Faggella (1:54)

That's going to be great. Diving into the world of skills training and basically exposing the trend that has been outdated for probably decades at this point. We've even seen Harvard and Deloitte basically come to the conclusion that the old model of sending everyone through the same course once a year is obviously no longer cutting it. And that's going to be the essence of our conversation today. And I'm going to head it off with asking you straightforwardly, what's the core thing enterprises have gotten wrong about how they've historically approached learning and development?

Nicola Lyons (2:24)

Ooh, where do we start? So I suppose the old model of enterprise learning has always assumed that we know what skills people need, we train them on them, we deliver that training, and then we move on. AI has made that cycle far too slow.

I think we've all had those experiences where our beloved and well-intentioned L&D colleagues have assigned us a load of video content on a MOOC and we're clicking through it, we're watching it at 2x speed, and then we don't retain anything in that training. And that's not efficient when AI skills and all skills are changing continuously in the age of AI. So historically, we've probably treated learning as more of an event, a once-off intervention that happens in a team where we identify the skills we need. We spend months designing or procuring a course to close that skills gap. We put people through the training. We measure their completion, check the box of whether they watched the videos or not, and then we move on. And that model was already very outdated and had its limitations, but AI has really exposed to those limitations. So the pace at which the technology is evolving, it's faster than any traditional curriculum development cycles. I've seen my team used to previously, a course would take three or four months to develop. Now they can turn out the same content in two weeks' time. And it's not a matter of the fact that we can create skills and content quickly.

It's more that we can personalize the learning content for people now. So we don't have to say we have 500 engineers, let's put all 500 engineers to AI training. We can acknowledge that all of those engineers don't have the same starting point. So a senior engineer might be leagues ahead of someone building edgantic systems and they don't need the same intervention as someone who's just downloaded and opening GitHub Copilot for the first time. We often don't have a sufficiently good baseline of actual capability either before training. So we know that a job title isn't capability, self-reported proficiency isn't capability, and neither is just completing a course, like we said, like checking the box. So that's particularly important because AI skills degradation and skill change are happening simultaneously. So some existing skills are becoming less valuable, some skills are becoming more important, particularly the cognitive skills, the creative skills, the element of the human in the loop, and some fundamentals are becoming more important now because you're responsible for evaluating work that you didn't actually produce. I think we've actually historically been too good at measuring learning activity and not very good at measuring actual capability change. So completion rates tell us one thing, they tell us that someone completed a course. They don't actually tell us whether someone is better at their job as a result of the training. That's the primary challenge that enterprise learning and development has had to date.