This subgroup of OGM spun out of a few impromptu Zoom calls with folks who place high value on permalinks, open data sharing, and other ways that our information Commons might be made more useful and accessible.
These calls are pretty geeky and take place every Wednesday. Ping Jerry if you’d like to be added.
on the limitations of the distributed-first approach, and whether they are inherent to the model or just tend to produce interfaces of a particular kind
it seems likely that [[nuggets]] that are intended to be used within a [[neobook]] are going to be even harder to integrate elsewhere once "enhanced" by AI
AI will tend to ‘overwrite’ towards a particular problem
[[Peter Kaminski]] would agree, and in the case we’re discussing the AI expansion made the end result harder to process. But AI in general seems like a powertool: it can be used with skill and make you go faster, but they can get out of control.
[[Flancian]] on maybe defining nugget as ‘human written’ or at least ‘full of human meaning/curated by a human’.
[[Chris Aldrich]] have you thought about going from Agora nodes to books/other format?
#AZS wish there were different models of federation
on being more focused on being broadcasters than collectors
#CA how do we map thousands of years of social evolution onto the possibilities of digital communities
[[the internet con]] by [[doctorow]] makes the case that we should go towards smaller distributed communities that can make their decisions
on how choosing an instance or platform is also about choosing what one wants to signal (or broadcast)
[[posse]] lets you pick and choose which communities you cross-post to; if you’re [[indieweb]] first you probably don’t care as much about the visibility that one single instance has over the rest of the fediverse
#AZS people are not necessarily posting on social media with the assumption that everybody will care; they assume there is in practice a small neighborhood. That is true until it isn’t (see e.g. unexpected virality)
#JM about [[boundaries]]: a well functioning group understands its boundaries and its norms
Expectations around where and when you expect people to engage with you.
Maybe we could carry our own ‘terms of service’ / privacy and engagement settings in a platform-agnostic way
#CA media has adapted and changed in its standards as well; newspapers in the early 20th century had completely different standards w.r.t. privacy expectations than towards the end.
Pete is proud of the timezone legend columns at the left
stopped updating in Sheets because some boxes and their borders span multiple cells; like overall look, but need a better grid editor that is better at multiple cell boxes
Also a useful characteristic to think of for both neobooks and open letters is portability
+1, and publishing has been a key factor in which open letters became popular/"successful" for a purpose in the past. MLK’s publishing story is interesting
[[aram]] on the challenges of independent tools in this space — in particular the problem of email deliverability (negotiating automated anti-spam false positives as it affects newsletters)
Assignment
Compose a 1 paragraph nugget on something that struck you in a [[FotL]] meeting
working on a demo of a media player and website configuration that lets you build static pages but have them act in a single page application style way, with the particularity that the media player is sticky/always on top while you browse around the site
#CA[[the great conversation]]: "the reason you want to read Aristotle and Plato is that they started a conversation we’re still having thousands of years later".
do we need a ‘media drumbeat’ to push back against this kind of thing?
where?
on the take "[[podcasts]] have become right wing media"
content policies are harder to enforce in audio land as it’s easy to distribute (unlike video) and not too easy to parse/consume/detect (unlike text) (for now?).
a space - the definition of an agora the one best known
public, mostly virtual, but has semi public sub-spaces (stoas). Different groups met in the different spaces and allowed people to join in and be part of something.
Go beyond the commons concern of the market
A distributed knowledge graph
Give it a list or repositories or resources and the agora will try to mash them together, find connections, find patterns, connect them.
A social network
an integration between social graphs and knowledge graphs and hopefully will make it part of the fediverse and t/f give it a clearer part of the social network concept.
Both a vision and a hypothesis
pro-social subsets of the internet would benefit from wide availability of a free [[interlay]] (as in the [[underlay project]]) provisioned and governed by a community as a commons.
For the common good.
Follow the principles of a commons.
Develop tools and instructions to make it more likely that such a connective layer arrises with these characteristics.
The agora is not good enough yet to be the connective layer but aims to be a bootstrap.
Note taking, wiki building, web annotating, communities, seem to have unique opportunity at hand. Possible to self-organize in cooperative groups and set up ways to federate within a commons.
Find common patterns and exploit them using the system where we can.
Design principles
Simple as possible
Leverage existing conventions and formats
bootstrap and build better ones from the bootstrap
Key characteristics
free software enabling a community to provision a basic knowledge commons
requires little of would be integrators and give back generously to participating communities.
inclusive and makes use of existing conventions, formats, tools, and networks for as long as practical.
Architecture
Agora Bridge
Software
User repositories
Social media
…
Imports users’ repositories every 30s and handles them. Usually git. Takes MassiveWiki and Social Media
Agora root repository
List of repositories
Configuration and bootstrap procedures
Base content
Agora Server
Software
Web interface
Interlink procedures
Accessed by browser
Python and Flask server
Points to a root repository and with those resources attempt to find the patterns and pull out nodes which then get served to the user.
May present notes on the same node by different users across different contexts.
Node
High level entity
Location on the knowledge graph
Can hold information from multiple users and join different files
Node can have Subnodes by different authors.
Links in Nodes will be seen and link those nodes together at the Agora level.
"chat-gpt, build me something like the ethereum network to distribute books"
we’re not anywhere near there yet.
"what is the next thing", in the sense of Einstein in 1904, is not something that this level of AI can solve.
[[JM]] unclear on whether LLMs can only regurgigate known things or can actually improve on them — compare with e.g. alphago coming up with novel moves
Took the full July off and did a family cruise to Alaska :)
Not as rainy as expected, gorgeous scenery
Q: did you keep a journal/notebook?
A: did, but haven’t done anything with it yet
[[Jerry Michalski]] One question that came up in a recent conversation: now that AI is around, do we need to still bother to take notes/curate links? :)
Yes :D
[[Chris Aldrich]] AI as it is now is interesting but it lacks:
In the late 1850’s, "Virchow was the first to correctly link the origin of cancers from otherwise normal cells, believing that cancer is caused by severe irritation in the tissues (the βchronic irritation theoryβ). Not all of his work was correct, however. He also proposed that cancer spreads around the body by the spread of the irritation in liquid form."
|
| Openness | How well does it play with other tools? How interoperable and customisable is it? | F/OSS. Use open standards. Has an API. Import/Export Allow 3rd party plugins. Available on all OSs, platforms. Supports structured data |
| Note-making | How good is it as a note-making tool? | Good outliner (easy indent/outdent, reorder, fold/zoom…) Atomic note structure, block references. Zettlekasten support. |
| Writing/Publishing | How good is it as a writing & publishing tool? | WYSIWYG, zen writing environment/design, editing feature set, seamless publishing system to Web, Social. |
| Idea discovery | Does it help users organise and (re)discover relevant notes? | Backlinks, Good search, AI-driven content discovery |
| User-friendliness | How easy is it to start benefiting from it, off the shelf? | New user can get started and get benefits without taking courses, reading 3 books and 23 blog posts, or installing a dozen plugins |
| Power | How powerful, adaptable & configurable is it? | Lots of features, highly customisable, many plug-ins, queryable knowledge, etc. |
| Multiplayer | Allows teamwork | Group-based collaboration: groups, permissions, workflows, version control, commenting, moderation… |
| Help | Is it easy to find high-quality documentation and answers to common questions? | Well-organised onboarding, documentation & FAQ system. |
| Community | Is there a supportive community around the tool | Helpful community, active 3rd party developers, etc. |
| Cost | How expensive is the tool | Free, or at least a v. useful free version in a freemium model |
| Data sovereignty | Who owns your data? | Your files on your PC. |
20:35
Jerry Michalski
Jerry Michalski says:BlueSky protocol
20:35
Aram Zucker-Scharff
Aram Zucker-Scharff says:oops wrong paste
Aram Zucker-Scharff says:
https://kolektiva.social/about
20:41
Jerry Michalski
Jerry Michalski says:small note: I’m Twitter user #509
Jerry Michalski says:when it was just an SMS service π
20:43
Aram Zucker-Scharff
Aram Zucker-Scharff says:π
20:43
me says:Pleroma
me says:Lemmy
me says:Misskey?
20:44
Aram Zucker-Scharff
Aram Zucker-Scharff says:
https://github.com/misskey-dev/misskey
20:45
me says:some interesting questions in this space:
me says:1. what is the biggest overall activitypub instance?
me says:2. what is the biggest which isn’t mastodon based?
20:46
Chris Aldrich
Chris Aldrich says:I think
mastodon.social
is the largest instance
20:46
me says:3. how large could a non-mastodon instance feasibly scale up to be currently?
20:46
Jerry Michalski
Jerry Michalski says:advertising is the colonizing force of capitalism in internet spaces
20:47
Aram Zucker-Scharff
Aram Zucker-Scharff says:Like
cohost.org
is sort of pushing in that direction as well
Aram Zucker-Scharff says:
https://pinboard.in/
also sort of monetizes that way
20:48
me says:
https://social.coop
20:48
Jerry Michalski
Jerry Michalski says:guilds —> unions
Jerry Michalski says:mutual aid —> cooperatives
20:50
Aram Zucker-Scharff
Aram Zucker-Scharff says:?
https://www.goodreads.com/book/show/4467218-travelling-brothers?from_search=true&from_srp=true&qid=wM2PoCC5zD&rank=12
?
20:51
me says:anti disintermediation
me says:crypto is a convoluted way to define a trust root (in most applications, it seems)
me says:"what are the platforms and protocols that we will take for granted 100 years from now"
me says:counter anti disintermediation π
me says:citation needed π
20:54
Jerry Michalski
Jerry Michalski says:full title: Travelling Brothers: The Six Centuries’ Road from Craft Fellowship to Trade Unionism (1979)
Jerry Michalski says:A Collective Bargain: Unions, Organizing, and the Fight for Democracy (2020) ?
20:55
me says:
flancia.org/mine/flanbook
is a bit about this — coopting the machinery of capitalism for the purpose of advancing the revolution
20:56
Jerry Michalski
Jerry Michalski says:oooo!
20:56
me says:commons
me says:enclosure of the commons == another manifestation of capitalist colonization
20:57
Jerry Michalski
Jerry Michalski says:my notes from this call:
https://bra.in/3pWnJo
20:58
Aram Zucker-Scharff
Aram Zucker-Scharff says:Ah i was thinking of Fight Like Hell by Kim Kelly
Aram Zucker-Scharff says:I’ll read that! :walking:
…but as long as these platforms provide this data in a documented way, do we need a central tool? wouldn’t collaboration emerge even without a central tool?
[[jerry]] +1 — the big fungus is a metaphor for the result of this collaboration. But [[playnz]] might be an [[mvp]]/[[demo]] or just an experiment/playground for interoperability.
In general though yes, the current implementation is Python-heavy as that’s what I use the most :) but we very much welcome alternate implementations. I also know that someone in the [[fellowship of the link]] was recently looking into adopting some Agora patterns (basically clustering relevant data from N sources into an aggregated node-like view) in a Rust-based project of their own.
Very nice, thank you! Today I finished work late and I’m now in the move, but I’ll try to integrate and test tomorrow before/after [[fellowship of the link]]