The Corporate Learning Roundtable: AI Is the Sidekick. We're Still the Lead Actors.

ANONDA roundtable invited straight-talking people in a room to a discourse around how humans are still in control.

EVENTS

ME-HR & Learning

10/2/20262 min read

Yesterday ANONDA hosted a roundtable on AI in learning and development. Sixteen people, most of them running L&D or people functions inside their own businesses. No panel, no script. Just a room of people trying to work out the same thing: how much of this job can you actually hand to AI, and how much of it was never up for grabs.

Dr Eli Bendet-Taicher opened with the line that set the tone for the next two hours: AI is the sidekick, not the lead actor. In a good film the sidekick makes the story sharper. In a bad one it steals the whole scene. The job of a good L&D team is knowing which film you're in.

THE REAL QUESTION ISN'T "CAN AI DO THIS"

Every AI conversation eventually gets stuck on capability. Can it write the course. Can it run the simulation. Can it personalise the content. The room moved past that fast, because the honest answer is yes, mostly, and it's getting better every quarter.

The harder question is which decisions you're willing to let happen without a human checking them, and which ones you still own the consequences of. That line moves. Governance rarely keeps up with it.

CONTENT DELIVERY IS A COMMODITY NOW

Alexandra asked the question every learning vendor should be losing sleep over: if a company's internal AI already knows an employee better than an outside provider ever could, what's actually left to sell?

Eli's answer was blunt. Pure content creation gets replaced. What survives is the consultant who understands the business goal before the training brief, defines what behaviour change actually looks like, and can prove it happened. Amy made the same point a different way: treat AI as a shift in how the operating model works, not a tool you bolt onto the side of it.

That's the split we saw across the whole room. The teams still asking "what course do we need" are exposed. The teams asking "what's the root cause, and does training even fix it" are the ones AI makes faster, not obsolete.

WHERE AI STILL HITS A WALL

The breakout groups landed on the same two things almost independently.

First, empathy and delivery under pressure. AI can generate the content. It can't read a room that's stretched thin, adjust on the fly, or carry someone through a hard conversation. That's still entirely a person's job.

Second, the mindset shift itself. More than one person compared learning AI to learning a language. You don't get fluent by reading the grammar book. You get fluent by getting it wrong in front of people, safely, until it stops being frightening. Most of the resistance in the room wasn't about the technology. It was about not having anywhere safe to fail with it yet.

WHERE THIS LEAVES YOU

If you're running L&D, watch out for using AI to answer the wrong question faster than you used to. Diagnose first. Correlate performance against the learning data you already have. Test small before you scale anything. Sometimes training is the answer. Sometimes it's one part of a bigger fix. Sometimes it's not the answer at all, and no amount of AI changes that.

Thanks to Eli for opening the conversation properly, and to everyone in the room who made it worth having.

INONDA runs these roundtables to get straight-talking people in a room without the panel and the pitch. If that's useful to you, get in touch.

Check out our upcoming events: https://www.inonda.co.uk/corporate-learning-table/register

Alex’s profile

linkedin.com/in/alexsardinha