In order to distinguish regular bots (wanted onwiki) and LLMs (unwanted onwiki), there should be a new icon for the latter. I propose a combination of the current robot icon and the pencil 'edit' icon in the bottom corner.
Description
Description
| Status | Subtype | Assigned | Task | ||
|---|---|---|---|---|---|
| Open | None | T141801 WikimediaUI theme (Codex/OOUI) icon inventory (tracking) | |||
| Open | None | T406077 [New component] Codex icons in message boxes | |||
| Open | Feature | Chaotic_Enby | T405137 [New component] Create a large-language model icon |
Event Timeline
Comment Actions
A few icons I've been having in mind:
Here's what they would look like at the "regular" icon size:
And here's the SVG file if anyone is interested:
Comment Actions
LLMs can be a useful tool if used correctly.
This paranoia about LLMs is a bit like the paranoia that a lot of people have that “OMG! Anyone can edit Wiki!”
Comment Actions
@Bugreporter2 This is not a forum discussion about the usage of LLMs. Please discuss this elsewhere.
Comment Actions
What's the current icon for regular bots? I think it would be better for LLMs to have the word "LLM" inside the text bubble. I think it would be more clear and avoid some confusion. Otherwise the reader could miss it, the elements used on the currently proposed icons are commonly used in other contexts unrelated to artificial intelligence (example).
Comment Actions
The current icon for regular bots is the robot face that is integrated here as a design component. The worry with "LLM" is that it isn't translingual, and we would need a translated version of the acronym in every language to avoid Anglophone bias, while icon combinations such as "robot" + "speaker" or "robot" + "text bubble" convey this more universally.
Comment Actions
The previously proposed icons don't strongly read as LLMs or generative AI to me. I also think if we go for something similar to my proposal we should avoid human-shaped brains or things that anthropomorphize AI as that would be misleading.
Made this quick concept by manually editing the code from File:OOjs UI icon robot.svg. It's not perfect, and it could probably get more simplified if needed / the neural network lines could be slightly thickened to display properly on smaller sizes, but you get the idea.
The previously proposed icons don't strongly read as LLMs or generative AI to me. I also think if we go for something similar to my proposal we should avoid human-shaped brains or things that anthropomorphize AI as that would be misleading.
Comment Actions
Been perfecting it so that it thumbnails better, I think this looks pretty good now, it would be neat to get feedback on it.
@Chaotic_Enby I've also seen File:WikiProject AI Cleanup (purple background).svg. At first glance it signals "unwanted bot" rather than "LLM or generative AI bot", which works too and I like how minimalist it is.
Comment Actions
The question that pops up in my mind is "do we have a situation where differentiating between AI vs LLM in software as a icon is useful?"
Comment Actions
@Soda, I don't think so, more reasonable ways of distinguishing them would be generative vs non-generative or automated ("bot"/"agent") vs human assistance. Also the level of autonomy, an OpenClaw agent and ClueBot NG are both technically "AI bots" but with very different expectations and capabilities (although that might fall under the "generative" aspect). Lots of constructive AI uses also have LLMs under the hood, and whether a small thing like BERT or a larger model is being used is less of a relevant consideration for the end user than what it does.