Dr. Poppy Crum

Or also even the environment.

If I'm trying to learn something auditorily and it's noisy, that's introducing extraneous cognitive load, right?

It's not the information itself, but it's because of everything else happening with that data.

And then the third is germane cognitive load.

And that's the load that is used in my brain to build mental schemas, to build, to organize that information, to really develop a representation of what that information is that I'm taking in.

And that germane cognitive load, that's the work, right?

And if you don't have germane cognitive load, you don't have learning, really.

And what they found is basically the germane cognitive load is what gets impacted most by using LLMs, which is, I mean, it's a very obvious thing.

Like that's...

Using LLMs, you're not engaging the mental effort to build cognitive schema, to build neural schemas and sort of the mental representation of the information that you can interact with it later and you have access to it later.

And this is really important because without that, you won't be as intelligent on that topic, that's for sure, down the road.

Let me give two examples.

I have a doctor, I have a lawyer.

And both of them use LLMs extensively for searches, say, or for building information.

In one case, it's for patient aggregation of patient data.

In another case, it's for history of case files.

And that is the GPS that's happening in those spaces because those are the tools that are quickly adopted.

Where you have someone that is maybe came from a different world, has learned that information, has gone and worked with data in a different way, worked their representation.

That information is going to be better extrapolation.

It's going to be better at generalization.