Ken Blackman

Explain the problem as you see it

I want a field to be visible until I've manually set it. The default is triage/inbox. Once I've set it, the field can be hidden.

Why is this a problem for you?

It's handy that a field can be visible or hidden depending on circumstances, but my use case doesn't allow for an option that doesn't require me to either hide it or show it by hand.

Suggest a solution

A new option for field visibility: hidden when value is not the default.

Explain the problem as you see it

ideas.tana.inc has become overwhelming and unwieldy. A major contributor is the proliferation of duplicates.

There is a "Consider voting for these similar ideas" but it goes by verbal rather than conceptual similarity.

Why is this a problem for you?

1. I can't tell if an idea has already been suggested.
2. I can't tell real voter response—a popular idea is more likely to have been suggested multiple times, splitting the vote.
3. Browsing the database in general is pretty hopeless.
4. Dialog about an idea, which can be very helpful in itself, is split among various incarnations.

Suggest a solution

It used to be prohibitively laborious to identify all the similar ideas, but now we have AI. So I suggest:

1. AI-enable the "Consider voting for these similar ideas" function to suggest conceptually similar rather than verbally similar ideas.
2. Use AI to do a pass on the current database to identify the same or similar ideas, and propose to the posters of the duplicates to take a look to see if they want to merge, or keep theirs separate.

Explain the problem as you see it

In my search results I see a bunch of similar nodes and need to distinguish them by where they are in the graph. This comes up a LOT for me. Examples:

• I need to make an instance of a term or concept the canonical one, and make the others references to it
• I need to add content to the "right" node based on context

etc.

Why is this a problem for you?

The only way I know of to do this currently is to click through each node one-by-one, just to see where it is.

Suggest a solution

A system field "breadcrumbs" or "path" that can be shown in search results. Would save me a ton of time.

Explain the problem as you see it

One of the most important things I do with my content is interact with it using AI (ask questions of it, converse with it, summarize, evaluate, etc.).

The problem in Tana is AI's small context window.

I want to interact with any node and all its descendants, of any size.

I've seen the video that shows how to hack around the small AI context window for a specific use case in Tana. But it's pretty fragile. IMO this is a universal enough need that Tana should have some kind of built-in support. I'd love for it to feel seamless, the way it does in other AI tools.

Why is this a problem for you?

Currently I'm inconveniently exporting large content to other AI tools such as Elephas just to interact with it in this way. Then convert and transfer the relevant results back to Tana.

Whenever I update or make a change in my content in Tana... new export, start over.

Suggest a solution

One possible solution is to implement MemGPT, which simulates an infinite context window using techniques originally developed for memory swapping in operating systems.

This would also allow for, say, an ongoing dialog of any length, without ever forgetting what was said earlier. That would be lovely!

This would allow me to do in Tana what I'm currently using several other apps for, and save me a ton of time and unnecessary thrash (transferring into and out of, format conversion, etc. etc.)