rri-204-4 (situated explanations)

[chrisdburr]

The following root file has been created:

• Title: Situated Explanations
• Module: Explainability (SAFE-D Principle)
• Skills Track: RRI
• Section: 4

Tasks

• [x] Review root file (Clau)
• [x] Answer comments (Chris)
• [x] Draft web version (Clau)
• [ ] Review web version (Chris)
• [x] Create slides (Chris)
• [ ] Review slides and draft script (Clau)

Link to file: https://github.com/alan-turing-institute/turing-commons/blob/drafts/drafts/rri-skillstrack/rri-modules/root-files/rri-204-4.md

[chrisdburr]

@ClauFischer: I'm cretaing this issue before the root file is 100% complete. There are a couple of reasons for this:

1. There is a possible section on social trust and explainability, which I was wondering if you'd be interested in authoring based on the PED chapter you wrote.
2. I'd appreciate general feedback about the overall scope of the section before I decide what else needs to be added or revised. I don't believe the section is that focused at the moment.

[ClauFischer]

• [ ] Look up resources for ethics of generative AI: https://github.com/alan-turing-institute/turing-commons/blob/673de90f4aff4b8c0d3fd28ba380d26cb2587d07/drafts/rri-skillstrack/rri-modules/root-files/rri-204-4.md?plain=1#L40

• [ ] Insert explanatory graphic on attention mechanisms for LLMs: https://github.com/alan-turing-institute/turing-commons/blob/673de90f4aff4b8c0d3fd28ba380d26cb2587d07/drafts/rri-skillstrack/rri-modules/root-files/rri-204-4.md?plain=1#L73

• [ ] I'm not sure if this section is the clearest but perhaps it's just the nature of the topic: https://github.com/alan-turing-institute/turing-commons/blob/673de90f4aff4b8c0d3fd28ba380d26cb2587d07/drafts/rri-skillstrack/rri-modules/root-files/rri-204-4.md?plain=1#L62

• [ ] I'm not sure I understand the relevance of the situated cognition admonition. https://github.com/alan-turing-institute/turing-commons/blob/82de42b600ae7951651688eccb064cf44ac47882/drafts/rri-skillstrack/rri-modules/root-files/rri-204-4.md?plain=1#L181

• [ ] Add section https://github.com/alan-turing-institute/turing-commons/blob/82de42b600ae7951651688eccb064cf44ac47882/drafts/rri-skillstrack/rri-modules/root-files/rri-204-4.md?plain=1#L199

[chrisdburr]

The situated cognition section is an optional one. If you think it is distracting, feel free to remove it.

Let's talk about the explanatory graphic for LLM. It's not a trivial task.