Tools for Thought Rocks: February 2022
videoyt[ 2/3 ] Presentations
[ 0/0 ] The future of personal computing
videoStart/end
00:01:00 -- 00:25:00
By
Rethinking the user environments across your entire personal computing domain
How personal computers help to think
How we represents things affects our capability to reason about it
Rethinking the OS
Central elements
items
Example
notes, todos, emails, podcast episodes
Items might reference other items
Possible some items might be transcluded into others
This allows using and manipulating them from the other item's views
Associated items
When items show together in a context, the system is able to connect them
Also, browsing contexts (eg. browsing paths)
Non-volatile workspaces
Separation of data from services and interfaces
Their transformations can be captured
Daily summaries, activity logs, ...
Create automations from the changes detected
Eg. user-defined notifications
Advantages
Might enable data sovereignty
Higer-level primitives for software development
Reduced duplication in third-party software development
Users would be able to move beyond the app
References
[ 0/0 ] Questions
Start/end
00:25:00 -- 00:40:00
How would other parts, like algorithms fit here?
Start
00:26:00
Algorithms specifically might be implemented as system-wide views
Thoughts about mobile devices
Start
00:28:40
Services might support interfaces they are not directly addressed, as UI's are other components
How are items different from FS files
Start
00:30:20
Items can be supplied dynamically by "item drives"
What new math would help this kind of ideas
Start
00:31:00
Question asker is working on mathematics for digital representation and asks which kind of tools would be useful
Work will continue on this
How to deal with services that is too complicated/not itemized
Start
00:35:10
Examples
Slack
...
Lean on "item drives"/metalayer to link elements that do not support functions like two-way-linking
Security
Start
00:38:10
How to deal with accumulating this many micro-pieces of code from a bunch of people
Probably isolation is a good path
OCAP
Capabilities
Using twitter to cultivate ideas
Start/end
00:41:10 -- 00:58:50
Questions end
01:07:00
By
End product
Principles
Consider using Twitter not as a "village café", but as a Memex, a web of interconnected ideas
Having it be "blocks" be short helps this
Problem
twitter is not prepared for this, there's no easy way to do this
An extension (Twemex) was created to solve this
Techniques
Show most liked tweets from random followed user
Show most liked tweets from user in the selected post
Show conversations between the reader and the selected user
Search for tweets published by the reader's followees
Search for tweets published by a specific user
Future plans
Search likes and bookmars
Profile notes
Curate multiple highlights
Reflections
Tools can shape behavior
Extensions let us reshape it
How to preserve this (twitter)?
Decentralize it?
Questions
Start/end
00:59:20 -- 01:07:00
Does this find really old tweets?
Start
00:59:20
Probably, the API used by the extension is different from the public API
On sustainability: how can we help support it?
Start
01:00:00
Community support or subscription
Can the fact that many people use the extension help do things that are not possible by the API?
Start
01:02:08
(Not explored directly, but...)
The app can do things that the API does not directly support with some "clever engineering" (like searching for most liked tweets)
Thoughts on how Twitter can help things
Start
01:05:06
Question
Prefers for Twitter to bring in the functionality or better to stay independent?
Answer
does not expend for Twitter team to care about it, as it's very niche
modeling, simulation, feedback systems
Start/end time
01:07:00 -- 01:24:15
Looks for inspiration on CAD <-> SIM -> FAB systems
Closest local maximum on software
[ Functional Reactive Programming ]
Issue
chaos theory
References
References [ Bret Victor ] 's work
Manipulation software design is hard
Most software is information software
Pet peeves
what does "information" mean here?
Questions
Start/end time
01:24:15 -- 01:49:04
Viewpoint on what are the common compontents of Tools for Thought from this point of view
Start/end time
01:24:35 -- 01:31:51
Things will take of when "we can run each other's systems on our own systems"
References OCAP model
Roam-style interface
In real-time
[ 01:30:00 ] Issues are produced in the interface between the real domain and the domain represented in the tools
Comment on downsides of simulations
Start/end time
01:31:51 -- 01:39:12
Question asker mentions differences between simulation styles (like game engines) and functional styles (where everything is "separated") and that they don't mesh together
How do current works and areas are related to this?
Start/end time
01:40:07 -- 01:44:58
There's a groundedness on [ Control Systems ] and that field on engineering
Thinking of making a tool?
Start/end time
01:44:58 -- 01:49:04
Still on the discovery stage, not thinking of making a tool yet
Considering the downsides of developing tools with a small team for a small audience
Thinking about developing a sort of Personal OS
Thinking about not using text as the grounding model for systme design
Looking for a system that can be boostrapped without textual processing