See Mara at work
Every claim shows its source. Every uncertainty is flagged.
A short walkthrough of Mara's grounded-response UX. Watch how she answers a research question — and what happens when she isn't sure.
This is a hand-crafted preview. Mara's grounding engine ships in the next release; the UX you're seeing is the target experience.
JT
You
I'm writing the introduction to a chapter on cognitive load theory. Can you summarize the original Sweller framework and its three load types, with sources I can cite?
M
Mara
Cognitive load theory was introduced by John Sweller in 1988, framing learning as constrained by the limited capacity of working memory 1. The framework identifies three distinct types of cognitive load:
- Intrinsic load — the inherent difficulty of the material itself, determined by element interactivity 2.
- Extraneous load — load imposed by the way information is presented, which instructional design can reduce 2.
- Germane load — load devoted to constructing schemas that aid future learning 3.
- Sweller, J. (1988). Cognitive load during problem solving: Effects on learning. Cognitive Science, 12(2), 257-285. doi:10.1207/s15516709cog1202_4
- Sweller, J., van Merriënboer, J.J.G., & Paas, F. (1998). Cognitive architecture and instructional design. Educational Psychology Review, 10(3), 251-296.
- Sweller, J. (2010). Element interactivity and intrinsic, extraneous, and germane cognitive load. Educational Psychology Review, 22(2), 123-138.
What just happened. Every factual claim in Mara's response carries a citation marker linked to a real, verifiable source. The grounding badge shows how many distinct sources were checked. Click any citation to see the snippet Mara pulled the claim from.
JT
You
Has the germane load construct held up empirically, or is there current pushback?
M
Mara
The construct has drawn meaningful pushback. Kalyuga (2011) argued germane load is not separately measurable from intrinsic load and proposed collapsing the framework to two categories 1. Sweller acknowledged the critique in 2010 and reframed germane load as "germane resources" rather than a third load type 2.
Mara is uncertain about how the field has reacted since 2020 — should I search current literature?
Grounded · 2 sources verified · 1 uncertainty flagged
- Kalyuga, S. (2011). Cognitive load theory: How many types of load does it really need? Educational Psychology Review, 23(1), 1-19.
- Sweller, J. (2010). Element interactivity and intrinsic, extraneous, and germane cognitive load. Educational Psychology Review, 22(2), 123-138.
This is the difference. Mara doesn't smooth over what she doesn't know. The gold "uncertain" pill flags any claim that falls outside her verified sources — and offers a one-click action to ground the answer against fresh search results. Other AI tools will confidently invent a 2023 meta-analysis here. Mara won't.