I’d like to suggest an enhancement for ordering the Tags displayed at the bottom of a page. Currently, Tags appear left to right, starting from the topmost folder (furthest group). However, I’d prefer to see (on the left) the Tags associated with the current page first, followed by tags from higher-level groups sequentially to the right.
Additionally, it would be helpful to have a toggle in the inspector pane to easily switch between viewing tags from the closest group to the furthest, or vice versa.
To me, it seems as if the tags are ordered alphabetically. And afaik, they don’t refer to a “current page”, but to the currently displayed document. Do I misunderstand what you’re saying?
Yes. I agree that Tags are ordered alphabetically. I am suggesting an order by proximity. So that I can at a glance determine if there are Tags set only for the current document and see a sort of a breadcrumb for nested Tags.
I suppose you are interested in finding tags that where used only in one document or a few, in order to replace them with tags that should have been used - well, what I do is look at the tag list itself, and spot the tags with 1,2… occurrences and investigate if the tag is good or was a fluke or repetition. I always find tags that shouldn’t be used.
If the AI is good enough to suggest tags with high accuracy, you may as well ask the AI directly to find you the document(s) you need. No tags needed.
It’s often envisaged that some kind of AI would enhance human-led organization. The more likely scenario, however, is AI replacing organization in its entirety. We are already seeing the later happening in e.g. the way social media curation algorithms work.
Tagging is a difficult subject. E.g. my tagging preferences change over time. Colleagues, family or friends wouldn’t tag like I do. AI can surely suggest tags but would it use the same tags that I would use right now?
Initially I was half-joking, but then I started thinking more about it. Not just tagging preferences can change over time (but this too can be AI’d, right?) - but tags actually can belong to different dimensions: subject matter; criticality, intended action, etc. In my case, e.g., I have tags such as “Botany”, “Critical”, “Review”, “Update”, “Thomas Artzt” (fictitious
Now, DT could look into the tags used by similar/related content the way you already do with classification, but it could also look at the content of a given record and infer whether you might want to tag it as “Critical”, or “Update”, based on these tags being used by other records AND having some (dimensions?) relatedness. It could be gentle about that, floating a panel of potential tags you could checkmark, and the panel itself would be off by default, or callable via a keystroke.
