Blog
L&D Saves AI From Being Generic
Marko Gargenta
Founder & CEO, PlusPlus
Intro
Last week I caught up with an HR leader at a large media conglomerate. Smart, experienced, rebuilding her learning organization from scratch. Ten minutes in, she asked the question I now hear on almost every call: will AI make learning and development irrelevant? Everyone can plug a question into a prompt. Learning teams are shrinking. What are we for?
Here is my answer, and it is the opposite of the fear. AI will not make the learning function irrelevant. It will make it the most strategic function in the company, because AI is only as good as the knowledge you feed it, and the knowledge worth feeding lives in your people's heads. The job of getting knowledge out of heads and into systems used to be called L&D. In the AI era it is called survival.
The tacit moat is the only moat
The reason internal AI initiatives disappoint is that the model knows everything outside your company and nothing inside it. It has read the whole internet. It can tell you, generically, how to manage a team, run onboarding, structure a performance review. That generic layer is the new table stakes, and every one of your competitors has the exact same layer.
What it cannot tell you is how performance reviews work at your company. Who your customers actually are. What your best engineer figured out last quarter that saved the release. That knowledge is tacit. Nobody wrote it down, because it was never anyone's job to.
The HR leader I spoke with put it better than I could. Expertise, she said, lives in people's heads, and it is very efficient right there. The expert applies what they know with no friction at all. Right up until the day they take a better offer. Then the intelligence walks out the door, and nothing and no one can access it again.
Generic AI plus generic knowledge equals generic output. The tacit knowledge in your people's heads is the only input your competitors cannot buy. That is the moat. The model was never the moat.
Your LMS knows who showed up, not what they know
The systems companies built to manage learning cannot capture any of this. A traditional LMS is an attendance machine. It knows who registered, who completed, what they rated the session. It knows everything about engagement and almost nothing about the knowledge itself, which sits locked inside opaque course files, authored top-down, by a central team, on a schedule.
So the org's official knowledge system holds the generic layer, the layer AI already made free. And the proprietary layer, the one thing worth capturing, has no system at all. It moves by hallway conversation and Slack thread, and it evaporates on the employee's last day.
Collective intelligence is the flywheel
The companies getting this right are inverting the flow. Less downloading content at people, more giving people a podium. Celebrate what individuals are building, automating, experimenting with. Let the expert record the five-minute video about the thing they just figured out. Then capture it, index it, and feed it into the AI, so the next person's question gets answered with your company's actual knowledge instead of the internet's average.
We call this collective intelligence: not artificial intelligence alone, but the intelligence of your people, continuously captured and fed back into the system. The more people share, the more others learn. The more that gets captured, the smarter your AI becomes about your world. A flywheel, and every turn widens the gap between your AI and the generic one.
Something else happens along the way. When people get a stage to show what they built, the fear goes away. Nobody is afraid of the machine they are teaching.
Which brings back the question from that call. Will AI make L&D irrelevant? Only the version of L&D that competes with the internet on generic content. The version that becomes the pipeline between what your people know and what your AI knows is not fighting for a seat at the table. It is the reason the table is worth sitting at.
The intelligence was never artificial. It was always in the building. The only question is whether it leaves in someone's head or stays in the system.
