Fellowship of the Link
-
a [[meeting]].
- Attending: [[x]], [[y]]
- Start recording, we’re currently using Jitsi and that works best with a [[Chromium]] based browser.
-
Threads from last time we would like to pick up:
-
#push [[fotl threads]]
-
[[SJ]]: missing you all :) Congrats to Jerry on the new site launch.
- Expanding a generative data commons w/ Mozilla + HF + Flickr; should connect w/ some of you about it
- Wikimedia is looking for a new ED, to start in January.
- US Wiki chapter is starting to do more portraiture @ events, via https://WikiPortraits.org — event suggestions welcome
- [[Shawn Murphy]] [[Flancian]] [[Aram Zucker-Scharff]]
-
Ahoy! Introductions :)
-
Shawn in Canada
- Aram in NYC
-
[[Hypergraphs]]
-
[[timeline]] visualizations (Aram’s work)
- https://context.center/timeline/covid/
- Q: [[http]] extension that aids archival/supports serving old versions?
-
Shawn: on versioning in general, attribution, provenance tracking.
- Would love to see the code!
- Aram: don’t remember this extension; am reminded of a DOM extension for offering e.g. archival copies for linked resources that can’t be retrieved anymore.
- Shawn: that seems like it could be a browser extension.
-
archive.org spec for time travel
-
Another web extension :) Annotations?
- Are they enough for a revolutionary change?
-
[[cognitive, evolutionary revolution]]
- Ahoy!
- [[Aram Zucker-Scharff]] [[Jerry Michalski]] [[Flancian]]
-
[[Jerry]] will be in Berlin next week!
- April has a speaking engagement
-
[[Aram]] is in high resolution today!
-
"Liberation day" (?)
- Because of the incoming tariffs
- Why did we let the right coopt ‘liberation’ and ‘revolution’?
-
Flancian: Can we use the Commons to fight back/build better?
-
[[Jerry]]: preparing something, a podcast associated with the substack.
- [[Aram]] nice throughput!
- [[Aram]] studies show that in some conditions glossiness of presentation can backfire; e.g. for vaccine information, some people believe "professional looking" sites less than amateur ones.
- [[Jerry]]: about an app called [[XyWrite]] XyWrite Wikipedia
-
Backlog
-
[[Jerry]] doing prioritization
- Dialing down sync meetings on [[Free Jerry’s Brain]], [[Neobooks]]
- But still working on Neobooks — trying to focus more on the deliverables.
-
On our "[[murder boards]]" (what a name): https://en.wikipedia.org/wiki/Murder_board
- What’s the ideal software?
- Something close to Zettelkasten but with a visual/2d component
- [[Jerry]]: a friend is doing https://sublime.app/
-
More stealth notes from [[SJ]]
- working on a Common Knowledge subset of the global knowledge index: all permissively-licensed works. https://docs.google.com/document/d/1X4Gy9fv53N8PUO5KgYFwL3Yo1TRtOQ1r6YjFiR4o14c/edit?tab=t.0
-
Could use help with:
- improving SOTA classifiers for topics; language; license
- improving graph-extractors for mapping documents to CC0 [[mesh]]es of claims. https://arxiv.org/abs/2502.19413
- improving QLever as best available upgrade for Wikidata backend
- divide-and-conquer implementing the above at web scale: developing a way to split up and do a bit of the processing at a time, incrementally updating the index.
- making and running benchmarks for the above
- also working on training better frontier-adjacent models, with Pleias, based on this data https://pleias.fr/
- helping IA Europe search for 1-2 technical staff for their Amsterdam office
- Not today! Stealth editing by [[SJ]] anyhow.
- Reminded that wikis need their own go-links and there is hot debate sometimes about who gets shortcuts for user-namespace pages :)
-
from the Build a Bigger Wiki Dept.
- Calling all indexers: [[SJ]] and [[Public AI Network]] folk working on building the index to a planetary library ~ https://publicai.network/atlas
- Also have new partners for [[Omnipedia]] in the Lam Lab at Stanford who want to feed STORM drafts into the draft-wiki. Experiments in the coming weeks…
- love to you all //
- Attending: [[Aram Zucker-Scharff]] [[Flancian]]
- Working while in vacation
- Training our replacements
-
Environment at the workplace
- Subscription woes due to owner statements
- Return to office five days a week
-
Personal projects
-
#AZS: [[foursquare]] space. There’s takeout data for the service that they shut down. Spent time to fill in checkins
- this could support/enable a social feature (through ActivityPub or AtProto)
-
on atproto vs activitypub
-
ActivityPub better for social networking itself
- [[betula]] example for non-microblogging functionality
- scaling is still a pain with AP
-
Atproto more sensible in some ways though
-
#AZS maybe nostr is the right technical answer, but the community has downsides
- and at this point social networks just decline
- Opportunity for "commons first solution that can make use of vault/archive data"
-
in "import/making data useful space":
- [[atproto]] conference this week
- Attending: <..>
-
[[Shawn Murphy]]: https://smurp.com/ presenting!
-
Q&A and comments
-
Very cool! Could nooviz be used as a library/embedded for graph/triples/[[TTL]] rendering?
-
Are you thinking of this as a hosted service or a toolkit that people can download and use?
-
Content security restrictions / iframe embedding
- Patched browser that user can specify embedding desired by user
-
We should meet!
- You should review the European Research Council: https://erc.europa.eu/
-
#PK
- perhaps related, the explanations in "A Thousand Brains: A New Theory of Intelligence", a book authored by Jeff Hawkins
- Hawkins says thought uses older brain mechanics based on physical maps of animals’ grid-based motion in (mostly 2D) space
-
#JM
- Attending: [[Chris Aldrich]] [[Jerry Michalski]] [[Peter Kaminski]] [[Flancian]]
-
Fires in [[Altadena]]
- [[Wind maps]]
-
Factors for these fires:
- Unusual driness
- Winds stronger than usual in this season
-
Building materials
- Could we build houses that can defend against fire?
-
Jerry: ice storm coming
- Had a very good speaking month so far!
-
Q: any feedback lately? How did it go?
- Started with talking about possible and impossible
-
Peter Kaminski
-
Two startups!
-
On the lowered barrier to entry to coding
- And regenerating codebases from improved prompts instead of (sometimes) maintaining the code
-
F: On the need to annotate generated content ideally going forward, traceability of meaningfulness (?)
- On losing touch with reality if we don’t do this
- The examples we already have of people not caring about factfulness
- Midterm elections
- Attending: [[Mathew Lowry]] [[Aram Zucker-Scharff]] [[Flancian]]
- Check ins
-
On [[atproto]]:
-
[[Mathew]] has been looking into it as it relates to possible plans/goals within the EU commission
- Each [[lexicon]] is a view or an [[app]] in the [[atmosphere]] ([[at-mosphere]]?)
- Thinking about how to make [[myhub]] an app in the atmosphere.
- Some of the content in the [[PDS]] could be published in a hub; some in [[Bluesky]].
- Thinking of e.g. managing notes in Obsidian, and publishing a subset of them through this pipeline.
-
About the question "Is [[Bluesky]] really decentralized"?
- A conversation that took place mostly in blog posts, but maybe could benefit from consolidating on a wiki.
- [[Mathew]] and [[Aram]] worked on a doc.
-
[[Aram Zucker-Scharff]]: a lot going on at work.
-
About the reaction to the new US status quo
- Surveillance of protesters started with the past administration
- The government is allowed to buy data from (commercial) data brokers, which in practice enables warrantless investigations (!)
-
Moving to Europe:
-
Check out [[archive box]]: https://archivebox.io/
-
Results from social.coop about defederating from threads:
Chris Aldrich can’t make it today. still dispdby fires and poor wifi conx.
- Attending: [[Mathew Lowry]] [[Aram Zucker-Scharff]] [[Jerry Michalski]] [[Eduardo Ivanec]]
-
Check ins
- Brain being rewritten for the Xth time
-
Idea: using AI to refactor/rewriting/making old code shinier
- Aram: it’s still far from being autonomous/decouplable. Some experiences with a python codebase it works great, little experiance with Rust as a codebase and it is useless to me.
- Flan: +1 totally.
- On resistance by engineering organizations, and the need for someone with domain expertise to review/validate.
- Mathew: it could work great for transpiling, updating, refining. A company could become an expert in applying genAI in partnership with those who understand the existing product and code.
- A lot of platforms / tools / profiles of labor will mutate with AI use
- On the kind of development processes this unlocks
-
[[FreeOurFeeds]]:
-
On Meta’s policy changes and fallout
- an off the record rehash of election 2024
- Attending: [[Jerry Michalski]] [[Peter Kaminski]] [[Aram Zucker-Scharff] [[Flancian]]
-
Check ins
- [[Jerry’s Reel]]: https://vimeo.com/1006299404
-
Peter’s [[AI coaching forum]] is now open
- Two paying members
-
[[Discourse]]:
- The [[AI Coaching Forum]] is Discourse based as planned, it turns out that there is a good official plugin to manage subscriptions
- An empty Discourse can be intimidating, but it has very good QoL once set up; and it’s super easy to install
-
[[Aram]] shares screen, [[context center]]:
- https://context.center/timeline/covid
- Based on Eleventy, Molly’s system before she moved to react, and Aram’s personal archive/pinboard
- It tries to produce archive links for every items
- Icons are new
-
Questions/feedback
-
Can we integrate this massivewiki, given that both use markdown as a source of truth?
- It would be nice to see a wiki in timeline view.
- How to set up one of your own?
- How to contribute new items or comment/engage?
- Long covid icon could be clearer; icons could surface the category on hover (use title property?)
- Love the deep linking + highlighting and archiving
- Maybe deep link the timeline and/or individual resources?
-
How are we doing with other projects we wanted to do this year?
Jerry says:The River vs The Village
20:45
CA
Chris Aldrich says:On being a secretary:
https://boffosocko.com/2015/11/02/on-being-a-secretary/
20:45
Peter Kaminski says:The River Where Black Swans Alight
20:45
Aram Zucker-Scharff says:
https://thepointmag.com/politics/the-bookmaker/
20:46
Chris Aldrich
Chris Aldrich says:For Flancian:
https://en.wikipedia.org/wiki/Satisficing
- Attending: [[Chris Aldrich]] [[Jerry Michalski]] [[Peter Kaminski]] [[Aram Zucker-Scharff] [[Flancian]]
-
Discussing the mind melding experiment
- Jerry did a new Brain export, sent over Mattermost
-
What have we been working on?
- [[Chris Aldrich]] enjoying the summer, working in a camp :) Plus jury duty
-
[[Peter Kaminski]] tried [[simplemind]] ~ [[mind mapping software]]:
-
[[Jerry Michalski]]
- lining ducks: website, reel
-
targetting [[2024-09-10]] to launch
-
writing [[wikily]], [[nuggets]]
-
Q: how to track authorship/state of editing in AI enabled flows?
- PK: for now I do it by hand by typing ‘this was written by <LLM> and not yet checked’
-
JM: wiki collaborative writing conventions could help here
- there was also a tool (Mixed Ink) that retained attribution
-
[[Flancian]]
-
[[Aram]]
- Working on shipping the [[timeline]]
-
Figuring out how to deal with different data exports
-
[[Flancian]] related to DSRs:
- https://x.com/exgenesis
- https://x.com/exgenesis/status/1825559564655497483
-
#AZS eleventy had a service that made it easy for people to self-host static sites based on twitter archives
-
is/was it https://twitter.11ty.dev/ ? They’re moving off Twitter to Mastodon, YouTube, Discord, GitHub
- very nice!
- would it be actually useful to offer this as a service? given that social media is so recency-biased
-
#CA reddits and forums also deal with the lack-of-wikiness (e.g. newcomers ask the same questions over and over again)
- it would be cool to go from ‘I’ve been writing things on twitter for years’ to ‘here is a narrative/a neobook/a story based on my writing’
- Attending: [[Peter Kaminski]] [[Jerry Michalski]] [[Flancian]]
-
[[Check ins]]
-
Jerry:
-
redesigning website
- reel is in production
- looking for speaking gigs
-
looking for ways to explain some concepts better:
-
Aram:
- trouble with audio in jitsi, at least on firefox
- happy with a new M1 based laptop, refurbished
-
using [[warp]]:
-
working on the [[timeline]] tool, iterating on moving elements to htmx
- could we use it within the fellowship?
- maybe extract links from these notes, and jerry’s brain entries associated with our calls, and put them in a timeline?
- #JM [[LATCH]]: Location, Alphabets, Time, Categories, Hierarchies
-
Peter:
-
Flancian:
-
Homework time
- We need a better word
- We want a bot to nudge us to do our homework halfway through the week (measured wed-to-tue?)
-
[[aram]] https://aramzs.xyz/essays/the-internet-is-a-series-of-webs/
- Aram looking at a series of articles about the future of the web being better than its past.
- Aram presented for five minutes, it was awesome!
-
[[semilink]] proposals
- Task: pick a link and do a writeup on it; then discuss for five minutes.
- Attending: [[Peter Kaminski]] [[Chris Aldrich]] [[Flancian]]
- [[Jitsi]] recording is on
-
Digital [[Syphons]] as an idea
- On the one hand it makes sense to keep copies of a lot of what we see
- But what about e.g. watching a movie? Would that give you the right to share the recording? It seems not (currently).
-
Chris on how the entertainment industry is shaped by this question and current events.
- [[Streaming]] as a factor
-
[[Piracy]]
- Example of a company recently shut down that was doing piracy on large scale as a service
-
#PK maybe we’re going into a future in which we’ll have:
- lots of derived work based on previous work, e.g. a million star wars
- interactive experiences
-
[[Mashups]] enhanced by AI
- Less creativity due to derivative nature
- But more creativity in the sheer audacity of the combinations possible?
-
#PK precedent in what happened once studio level recording/editing became widely available
-
AI derived work maybe doesn’t have the ‘cultural commonality’ sapect of current artifacts
- #PK but we can expect a few creators to be ‘outliers’, widely followed, providing commonality
- Influencers
- Explosion in cultural context may result in context collapse — or just something different
- Ingroups / see communities around influencers with their own memes
- #PK yet AIs could also be guides to such communities
- [[Extension 765]]
-
#PK from an org I’m part of, something to consider if you’re in California: "Please sign these letters to legislators, telling them that misguided AI laws will hurt startups and small companies and discourage AI innovation and investment in California." https://docs.google.com/forms/d/e/1FAIpQLSeR5VrXxDJA3sJtkWDAKLH1TT0havDxmCf9PYAupxECu1BQYw/viewform
- Oligarchy -> where the rich get to write the rules?
- #CA "The factory cannot only look at the profit index. It must distribute wealth, culture, services, democracy. I think factory for man, not man for factory, right? The divisions between capital and labour, industry and agriculture, production and culture must be overcome. Sometimes, when I work late I see the lights of the workers working double shifts, the clerks, the engineers, and I feel like going to pay my respects." — [[Adriano Olivetti]]
- Typewriters and degrading industrial quality, planned obsolescence
- #PK The [[Matthew effect]] of accumulated advantage, sometimes called the Matthew principle, is the tendency of individuals to accrue social or economic success in proportion to their initial level of popularity, friends, and wealth. It is sometimes summarized by the adage or platitude "the rich get richer and the poor get poorer". The term was coined by sociologists Robert K. Merton and Harriet Zuckerman in 1968 and takes its name from the Parable of the Talents in the biblical Gospel of Matthew. https://en.wikipedia.org/wiki/Matthew_effect
- https://en.wikipedia.org/wiki/Lindy_effect as it applies to typewriter
-
#F on the form earlier: it would be nicer to have something richer that lets you both support and oppose arbitrary proposals.
-
#AZS on the US trending towards maximum craziness
- Things have been crazy for a while to the extent that someone shooting Trump didn’t raise the bar that much
- On trying to, and failing, to keep up — like in the early days of the Trump administration
- The half-life of news seems shorter and shorter
- Have been building a full [[timeline of Covid]]
- On examples of what people/communities were able to build in the US with a glimpse of UBI and some housing stability
- #F on how we could learn more from past information
- #AZS [[Cory Doctorow]] gave a great talk at [[HOPE]]
- Attending: [[Peter Kaminski]] [[Flancian]]
-
[[AI Coaching Forum]]
- will be [[Discourse]] based
- would love a demo once it’s ready
- there will be a paywall
-
on the [[productivity equation]] of AI:
- AI gives a significant productivity boost
- so it should make it easy to give away a good fraction of content
-
lots of recent but by now historical data — e.g. how to best prompt a one year old model
- could make for good content to publish (on a delayed schedule)
-
the [[newspaper paywall]] model could also apply/have advantages
- free articles and gifting the obligation of signing up to read them
-
#PK Good post by Yoshua Bengio, sections of the post include:
- For those who think AGI and ASI are impossible or are centuries in the future
- For those who think AGI is possible but only in many decades
- For those who think that we may reach AGI but not ASI
- For those who think that AGI and ASI will be kind to us
- For those who think that corporations will only design well-behaving AIs and existing laws are sufficient
- For those who think that we should accelerate AI capabilities research and not delay benefits of AGI
- For those concerned that talking about catastrophic risks will hurt efforts to mitigate short-term human-rights issues with AI
- For those concerned with the US-China cold war
- For those who think that international treaties will not work
- For those who think the genie is out of the bottle and we should just let go and avoid regulation
- For those who think that open-source AGI code and weights are the solution
- For those who think worrying about AGI is falling for Pascal’s wager
- [[Reasoning through arguments against taking AI safety seriously]]: https://yoshuabengio.org/2024/07/09/reasoning-through-arguments-against-taking-ai-safety-seriously/
-
#PK Ray Kurzweil, inventor / futurist / AI optimist, envisions that humans will merge with AI, which may excite you or disgust you. He has a new sequel book, "[[The Singularity Is Nearer]]: When We Merge with Computers", and he’s doing interviews to promote the book.
- #F on identifying with AI and the steps/slopes that way
- "My (Unfiltered) Take on AI Safety" https://daveshap.substack.com/p/my-unfiltered-take-on-ai-safety
-
[[itbwtcl]]: "Sophisticated people deride Disneyesque entertainments as pat and saccharine, but, hey, if the result of that is to instill basically warm and sympathetic reflexes, at a preverbal level, into hundreds of millions of unlettered media-steepers, then how bad can it be? We killed a lobster in our kitchen last night and my daughter cried for an hour. The Japanese, who used to be just about the fiercest people on earth, have become infatuated with cuddly adorable cartoon characters. My own family—the people I know best—is divided about evenly between people who will probably read this essay and people who almost certainly won’t, and I can’t say for sure that one group is necessarily warmer, happier, or better-adjusted than the other."
- whole hog and drills
- "emacs outshines…"
-
#PK PDP-8 and similar experiences
- [[hackmd]] conflicts
-
[[bluesky agora support]]
-
[[firehose story]]
- Attending: [[Peter Kaminski]] [[Bentley Davis]] [[Flancian]]
- Green screens and [[Jitsi]] < [[Meet]] < [[Zoom]]
-
[[Flancian]]
- Introspection and soul searching :)
-
[[Peter Kaminski]]
- Doing well and making progress on how to deliver educational content
-
Epiphany with GenAI
- When "most" people can generate e.g. images via GenAI, where does that leave the creator/curator/etc.?
- The understanding of the craft is still valuable; the role might be helping people make better images/work with creative tools better
-
How will niches evolve with these changes?
- Some people are very good at writing prompts
- Some people have very good ideas
- Some people are very good at ‘finishing’/touching up
- [[movies]] <-> [[games]] spectrum
-
[[David]], CEO of [[Midjourney]]:
- [[multi-level]] architectures to develop full length books; difference betwen published fiction and interactive text games blurring
-
[[PKM]] and [[AI]] interactions
-
Interesting question the first
- Removing distinction between finding a note and writing it on the fly
- Tracking provenance
-
[[Color of the bits]]
- [[Marc-Antoine Parent]] - knowing the "edges" of a dataset, and knowing if/when the AI is reaching "outside" of the edges
-
Interesting question the second
-
Corpus gathering
- Original source data
-
Synthetic data
- [[Nemotron-4 340B]], an open-source pipeline for generating synthetic data. 98 percent of the training data used to fine-tune the Instruct model is synthetic and was created using Nvidia’s pipeline.
- Communities of practice and the potential of their corpora
-
[[Prompt router]]?
- #PK doesn’t use them (do they exist yet?) because the top-of-the-line models (e.g. [[gpt 4o]] currently) are good enough in most aspects
- #F But what if other models were optimistically/lazily surfaced? :)
- It is true that sometimes you know that your model isn’t doing as good as it could do (e.g. with coding)
-
[[LMStudio]]
-
Do we worry about enshittification in the AI space, e.g. [[OpenAI]]?
- Will they be captured by capitalism?
- Carrying the ring of power — why not have a fellowship?
- Why be so proprietary and non-open?
-
Do we need better [[open letters]]? ;)
- Attending: [[Jerry Michalski]] [[Chris Aldrich]] Late: [[Aram Zucker-Scharff]] [[Flancian]]
- (in media res)
-
On [[Jerry’s keynote]]:
- Modelled after [[April Rinne]]‘s earlier.
-
Keywords:
- [[design from trust]] — what we can unlock if we stop assuming everybody is out for the worst.
- four-s methodology <- focus of the second half of the presentation.
- About strategic thinking methods.
-
[[Chris Aldrich]] Ed DeBono (sp?) thinking hats for the new millenium — related to the preaching circuit.
- Friday appointment :D
-
On being in the [[speaker circuit]] / getting back to the arena.
-
Book question
-
What are book-equivalent projects in your past?
- Binders of research
- Newsletter production
- Brain
-
[[Chris Aldrich]]: on the lifecycle of ideas. Sometimes one outgrows ideas and they drop out of conversation because of this.
- [[Design from Trust]]: https://wiki.openglobalmind.com/projects/jerry’s_nuggets/design_from_trust_(neobook)
-
#AZS
- need to establish authority and standing
- need a hook into people’s attention (paraphrasing)
- the "easiest" (or most common and socially accepted currently) way of solving both is publishing a book
- respect for the [[POSSE]] but unsure it works for this specific problem — except for [[Cory Doctorow]]. But even he didn’t start that way. It is a solution intended for online identity as an object.
-
[[Jerry Michalski]] something that leads people to discover the thing themselves, so they want my help with it
-
#AZS
- I sorta think the answer lies somewhere in documenting the history of the thing
-
[[Chris Aldrich]] parallels between the publishing industry and the studio system, which is mostly owned by corporations for which making films is a secondary concern (e.g. Sony is an electronics company first).
- One of the questions that publishing houses ask is: ‘who already cares about you’.
- Attending: [[Jerry Michalski]] [[Aram Zucker-Scharff]] [[Peter Kaminski]] [[Flancian]] [[Aram Zucker-Scharff]]
-
Check ins
-
[[Flancian]] under a pile of work somehow! Will stay only for part of this unfortunately. But fine otherwise!
-
[[Jerry]]
-
[[Peter]]
-
[[ChatGPT]] and memory
- initially thought it was an "anti-feature"
- "could you keep separate memories for different things"
- asked them what they would call the categories around which memories were clustered
- they suggested ‘project’ or ‘topic’
- queried ‘books that came up in calls’, now interactively asked them to add the cyberfeminist index to a project
-
[[tools for thought]] in the age of [[conversational ai]]
- [[assistants]] participating in meetings and facilitating them
- [[shared memory in the commons]]
-
#AZS on the risk of enshittification of chat-gpt and it turning further into a walled garden
-
[[Chris Aldrich]]
- idea attribution and [[my self told my self]] :D
-
#AZS keyword: [[rendition]] as used in [[the age of surveillance capitalism]]:
- “The prediction imperative transforms the things that we have into things that have us in order that it might render the range and richness of our world, our homes, and our bodies as behaving objects for its calculations and fabrications on the path to profit….there can be rendition without surveillance capitalism."
- [[meta]] on the problem of search/retrieval
-
#F [[subconscious]] is gone/done :(
-
#PK fwiw, i don’t have many books in my chatgpt memory yet, so the Cyberfeminism book had big influence. i asked chatgpt about more books that would fit with others, and it listed these:
- "The Fourth Industrial Revolution" by Klaus Schwab
- "Technofeminism" by Judy Wajcman
- "Data Feminism" by Catherine D’Ignazio and Lauren F. Klein
- "Weapons of Math Destruction" by Cathy O’Neil
- "Algorithms of Oppression: How Search Engines Reinforce Racism" by Safiya Umoja Noble
- "The Second Self: Computers and the Human Spirit" by Sherry Turkle
- "The Age of Surveillance Capitalism" by Shoshana Zuboff
- "How We Became Posthuman: Virtual Bodies in Cybernetics, Literature, and Informatics" by N. Katherine Hayles
- "Emergent Strategy: Shaping Change, Changing Worlds" by Adrienne Maree Brown
- "The Future of Another Timeline" by Annalee Newitz
- Peter Kaminski
Peter Kaminski says:Borges’s Library of Babel has EVEN MORE books than the Library of Congress!
Jerry says:that’s the cheaper version
Chris Aldrich says:"No piece of information is superior to any other. Power lies in having them all on file and then finding the connections. There are always connections; you have only to want to find them."—Umberto Eco (Foucault’s Pendulum)
Aram Zucker-Scharff says:
https://github.com/AramZS/aramzs.xyz
https://www.flavorwire.com/515783/brooklyn-author-recreates-borges-library-of-babel-as-infinite-website
-
On scraping [[kindle]] / exporting our data from there
-
[[open letters]]
-
#PK like the idea but probably won’t work due to people finding it too challenging
- an alternate approach: make things that [[commonize]] stuff; focus on improving usability of tools aligned with our goals / the good of the commons.
- on manual indexing
-
[[semilink]] update
- Attending: [[Jerry Michalski]] [[Aram Zucker-Scharff]] [[Peter Kaminski]] [[Samuel Klein]] [[Flancian]] [[Aram Zucker-Scharff]]
-
On [[Google I/O]] and [[OpenAI]]‘s presentation
-
On [[wealth redistribution]] and giving back to creators
-
#AZS
-
on [[maximizing shareholder value]]:
-
Google Books
-
The Long Next
- [[SJ]] we don’t need to ‘fix capitalism’
- SJ says:most of society still runs on other isms
- me says:+1, we want to upgrade existing systems and build more and better bridges between them
- /now and ending urls with
/json
and /md
to get JSON and MD versions of the data respectively. Standard data sharing flows.
- Webfinger
- Webmention
- ActivityPub
-
https://github.com/jhass/nodeinfo
-
Ideas for shared projects
- Frame the fellowship as a community of practice, use knowledge tools together and showcase/present the results
- Organize classes in digital knowledge tools
- Take all our notes and artifacts and feed them to an LLM
-
Run a semilink (seminar) series
- One link 15m session - 3m, about link, 5m about discussion, then a wrapup. Embed the link in the full history of its context.
-
Let’s come back next week with ideas in any form you want and present
- We’ll call it [[semilink]] as per Samuel’s
- everyone has to bring one half-baked link and one potential person to include 😉
- Attending: [[Jerry Michalski]] [[Peter Kaminski]] [[Flancian]] [[Chris Aldrich]]
-
Daylight savings excitement :)
- [[flancian]] now moved the next two instances to an earlier timeslot in EU
-
[[check ins]]
-
Where do we want to go from here?
- Last time we discussed [[composer]]
-
[[Neobooks]]
- [[Flancian]] is a huge fan of [[silverbullet]]
-
[[Peter Kaminski]] working on [[Midjourney]] and [[LLM]] classes
-
#CA Notes from 1w-2w ago when we were talking about [[neobooks]] and a possible pipeline
-
having a repository of ideas
- generally indexed or interlinked
- having a tool for creating an outline
- having a tool to go from outline to full text
-
similar preexisting tools in this space:
- tools that take URLs and create ‘books’ for people to read as pdf or mobi
- #CA Here’s some of the prior art research on these sorts of reading lists: https://indieweb.org/reading_list
- #FL would like to experiment with this last process in particular (outline to full text, set of nodes to text)
- #PK A fast, free image viewer I use for sifting through thousands of images: Phoenix Slides https://blyt.net/phxslides/
- Attending: [[Peter Kaminski]] [[Flancian]] [[Samuel Klein]] [[Chris Aldrich]] [[Aram Zucker-Scharff]]
-
On [[neobooks]]
- The [[BFDL]] concept as it applies to this project :)
- Reference [[neobook]]/first [[neobook]]
-
[[nuggetization]]
- how it relates to translating between [[worldviews]]
- The difficulty of finding editors to work with to help work them into neobooks
-
[[Aram Zucker-Scharff]]
- 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?
-
Precedents for [[nuggetization]] and [[composition]]
-
composition:
- as per the above examples
-
nuggetization
-
Repository of good nuggets?
-
[[indie search]]:
- Attending: [[Jerry Michalski]] [[Chris Aldrich]] [[Peter Kaminski]] [[Bentley Davis]] [[Aram Zucker-Scharff]]
-
[[check ins]]
-
Next Monday: [[free Jerry’s brain]] demo, [[1PM Pacific]]
-
Topics from last time
-
[[collective authoring]] as distinct from the basics of [[book publishing]]
- the default sort of social interaction so far for neobooks are github/git forge fork-and-send-PRs
-
are there any less geeky ways to contribute/interact?
-
do we need an analog for copy-and-paste for how to collaborate in a distributed mesh of documents?
- there is a reason we end up with git as a backend; the alternatives are more computationally expensive, require servers or parties to run software.
-
[[google docs]] as a model maybe — the comments-only access is something we might want an equivalent for
-
on the topic of group memory; how not to start every conversation with a relatively blank slate
-
[[cross posting]] of [[bluesky]] and the [[federation]]
- https://snarfed.org/2024-01-21_moderate-people-not-code
- https://snarfed.org/2024-02-05_52058
-
#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
-
#CA [[Book Club]] on [[Cataloging the World]] and Index, A History of the
-
[[Intimacy Gradient]]
- 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.
-
On evolving [[civility standards]]
- [[Trump]] exploiting this
- Need different spaces and contexts where it one can be in rough notes, drafty, professional, etc…
- Attending: [[Jerry Michalski]] [[Flancian]] [[Chris Aldrich]] [[Michael Grossman]] [[Bentley Davis]] [[Aram Zucker-Scharff]] [[Peter Kaminski]]
-
[[Jerry Michalski]] and [[Peter Kaminski]] on a productive call about [[massive wiki]] and lots more!
-
Could we have a dashboard of all [[calls]] in our space
-
#PK has something Google Sheets based, using Sheets as a grid/table editor
-
Present as a neobook?
-
[[Flancian]]
-
[[Jerry Michalski]]
- How is it different from a viral post
-
[[Flancian]]
- Intention
- Notion of who is reading
- Some level of meta awareness
- There are some similarities
-
Neobooks and how they connect
- Same shape as an Agora node
- Open letters are hugely interesting.
- They get amplified with history
- Similarly new media
- The internet evolves similar
-
[[Aram Zucker-Scharff]]
-
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
-
[[Jerry Michalski]]
- composability
- https://bra.in/3jgE2p
- Manifestos - too many authors have robbed them of a degree of their potency?
-
[[Aram]] an interesting characteristic of open letters:
- in modern times open letters are implied to have many authors, or have been created with a degree of anonymity
- [[letter on corpulence]] :)
- on expanding circles of trust
- [[google docs]] as defacto publishing platform
-
#JM about inspirations for [[neobooks]]
-
which direction to go to now?
-
[[nugget a week]]
-
[[webmentions]] would be ideal
-
[[bridgy fed]] by [[snarfed]]
- some concerns about the difficulty of self-hosting
- for now some consider it a SPOF
- ideal model could be: something that works like nostr, where you can just point your system to a set of bridgy fed instances and it balances
- we can start by posting to the mattermost challenge — links to anything
-
[[IETF]]: a protocol is not ready until you have two independent implementations
Assignment
- Compose a 1 paragraph nugget on something that struck you in a [[FotL]] meeting
-
Attending: [[Jerry Michalski]], [[Flancian]], [[Aram Zucker-Scharff]], [[Peter Kaminski]]
-
Check ins
-
Useful open source video editing with
-
[[flancian]] presentation (maybe): https://anagora.org/go/agora-slides
-
three facets
-
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.
- Nodes are social
- See anagora.org/nodes for visualization
-
Common patterns:
-
You can agree on a slice for a concept and then right away start coordinating - exactly like using a hashtag
- Socail media works with hashtags to integrate into this concept.
- Agora lets you hashtag anywhere
-
Stoas
- Social too
- Locaitons that act as semi public spaces
- Set up a hedge doc that is totally anonymous.
- Will get processed into the Angora within 30 sec.
- Etherpad and Jitsi both can get pulled in to a Stoas.
- Heavily uses iFrames whenever possible
- Default empty pages are a call to action
-
How to join!
- Take notes in your fav tool
- Publish in platform - usually git.
- Let an Agora maintainer know about it.
-
Wikilinks everywhere
- Edges are [[wikilinks]] which refer to other nodes as found in volunteered subnodes
-
Idea of composing notes
- Some notes have special behavior
- [[go]] notes -
#go url
or [[go/foo]]
- [[pull]] - bring in the resource at bar with
#pull bar
- [[push]] - push out the resource to foo with
#push foo
- Attending: [[Jerry Michalski]], [[Chris Aldrich]], [[Aram Zucker-Scharff]], [[David Pickrell]], [[Flancian]], [[Michael Grossman]]
-
Introductions
-
[[Aram Zucker-Scharff]]
- Lead engineering for years on [[Pressforward]]
- tool making experience as it interacts with the indyweb
- current project: how we link note tools together. personally using [[eleventy]] based [[context.center]]
-
[[Flancian]]
- welcome [[David]]!
- SRE by profession here, knowledge commons/graphs tinkerer by hobby :)
- [[agora]] ~ https://anagora.org is my main project in this space
-
[[Michael Grossman]]
-
[[David Pickrell]]
- itinerant "software salesman" :)
- focused on helping people solve problems with software
-
one pitch for the group: son-in-law works on book.io which is a publishing environment that tries to empower authors
- unique covers for each of the books (NFT space, in the cardano chain)
- they release unique digital books at 20-30 USD a piece
- moving to do audio
- long time [[brain]] user
-
What’s on people’s minds?
-
[[Michael Grossman]] curious to ask David about book.io — there are public domain works with many covers, how does that work?
- artists working with GenAI
-
[[Jerry Michalski]] is book.io a scarcity play or a publisher which is trying to be useful to creators?
- it tries to be the second
- targets mostly self-publishing authors currently
- every time a book gets resold, the author gets a percentage
-
[[Aram Zucker-Scharff]]
- write professionally, would like to use editors more, and would not like to use them for free.
- would like to use such a platform to give participating editors a cut of produced works.
-
[[Chris Aldrich]]
-
reviewed some books in book.io
- there is no translator information for translations of public domain books, e.g. [[anna karenina]]
-
Presentation on the Agora?
- 20th of September — the date is set! :D
-
Fediverse and Agora integration in the works
- note taking in the fediverse: exploring this space
-
[[AZS]] don’t know anything in this space.
- obsidian can push to the agora, then the agora can provide activitypub services
- [[Chris Aldrich]] there is potential for UI innovation in this space, something beyond streams
- wikilinks everywhere :D
-
[[does chatgpt obsolete notetaking]]
-
[[AZS]] typical AI-disruption-take.
- These are tools for thinking and not just for ingestion.
- our computers/mobile phones already turn us into cyborgs; AI just adds another layer to this
- human intervention is what gives these things/outputs a value
-
[[CA]] note taking is a ratchet
-
"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
-
[[Michael Grossman]] podcast: https://a16z.simplecast.com/episodes/a-true-second-brain-xrODaBD2
- on [[Nat Eliason]]
- being able to take notes over years and put them into a model which you can then query/search
- a certain type of note taking will be much more useful than in the old days
- surfacing the note that somebody you trust has shared with you is what’s golden
-
[[Jerry Michalski]] when Google came out, we outsourced our memories to Google
-
[[David Pickrell]] recording/transcription services you can add to meetings.
- using [[ChatGPT]] can help get a sense of material but might not help actually learn that material
- [[JM]] [[story threads]]
- [[AZS]] wikipedia is essentially the biggest collection of knowledge, yet people also get degrees
-
chat / link dump
|
| 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
20:36
Jerry Michalski
Jerry Michalski says:
https://blueskyweb.xyz/blog/10-18-2022-the-at-protocol
20:37
me says:
ctzn.one
— now down probably
20:37
Jerry Michalski
Jerry Michalski says:Paul Frazee?
20:38
me says:yes:
http://anagora.org/pfrazee
20:39
Aram Zucker-Scharff
Aram Zucker-Scharff says:
https://paulfrazee.medium.com/back-to-basics-what-is-the-point-of-decentralization-a495123a1fd7
Aram Zucker-Scharff says:
https://context.center/topics/cryptocurrency/#activity-pub
20:41
Jerry Michalski
Jerry Michalski says:small note: I’m Twitter user # 509
Jerry Michalski says:when it was just an SMS service 😃
20:43
Aram Zucker-Scharff
Aram Zucker-Scharff says:😃
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
Aram Zucker-Scharff says:
https://twitter.com/Chronotope/status/1587511147137486848
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 😉
me says:citation needed 😉
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:
2022-06-24
Video
https://www.youtube.com/watch?v=K9aRBtTqLjw
Notes
-
Discussion of [[go links]]
-
Time?
-
[[shared resources]]
-
[[Chris Aldrich]] putting up something in the site
- here’s what we’re doing, here’s where we’re going, here is how to join
- let’s use [[chat]] for coordinating
-
what’s a high leverage topic for the three of us?
-
[[Jerry Michalski]]
-
from chat unprocesssed
- Jerry Michalski dice:transclusion
- Chris Aldrich dice:How to create a Read Fork Write Merge Web?
2022-06-17
2022-06-09
Matterbridge site for group communication: https://chat.collectivesensecommons.org/agora/tips
https://collectivesensecommons.org/
Mathew Lowry mentioned [[Mysilio]]
[[David McRaney]] (twitter) has several books:
2022-06-02
- Attending: [[Jerry Michalski]], [[diego de la hera]], [[flancian]], [[Chris Aldrich]]
-
Our personal history
- Diego en Córdoba
- Jerry Michalski pasos por sudamérica, ahora en Portland
- Eduardo de Buenos Aires, ahora en Suiza
- Relationships locally and globally
-
Jerry: [[open global mind]] meeting just before
-
([[Chris Aldrich]] joined, introductions)
-
[[Jerry Michalski]]
- [[the brain]]
- [[chris]] any shares in the company?
- not anymore
- do you have backups?
-
occasionally :) but the format is a big bag of [[json]] objects, not bery interoperable
-
[[diego de la hera]]
-
[[Chris Aldrich]]
- background: biomedical and electrical engineer, been on the web since the early 90s
- [[indieweb]] / https://indieweb.org
- [[open web backup]], strong belief in owning your own domain, your own data, your own identity
- done work on genetics, but germane: spent a lot of time in the [[entertainment industry]] as it overlaps with the open web
- massive amount of data
- history, culture as intertwined with knowledge tools
- current focus on wikis and note taking tools
-
[[Jerry Michalski]] wishful thinking for a design in this space
- it should allow all users to view data through their chosen tools
- sometimes you change tools as you shift what you need to do (they are good at different modes of thinking)
-
[[planes]]
-
[[kumu]] == https://kumu.io
- [[jupyter notebooks]] are an inspiration
-
https://thebigfungus.org
-
[[diego de la hera]] we mentioned a diversity of tools here:
- closed and open
- wiki like and otherwise
- …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.
-
are these tools perhaps already playns?
-
[[Chris Aldrich]] [[progressive enhancement]] is key
-
[[flancian]] problems and opportunities:
- exploring hub architecture vs full graph integration
- lossiness as you interop through tools, playns
-
complexity of integrations
-
[[translation as a service]]
- [[markdown as lingua franca]]
-
[[Jerry Michalski]]
-
interest in the [[crystallization]] of what we know together
- [[fellowship of the link]] as it relates to other projects
-
[[diego de la hera]] isn’t the web as a whole what we need? :)
-
[[lists]] we could assemble:
-
people in this space
- the above
- [[tantek Çelik]]
- "how can we build this with as little technology as possible"
- projects in this space (knowledge tools)
-
recurrent calls in this space
- possible [[lingua francas]]
-
[[Chris Aldrich]] "The world lost so much because Vannevar Bush didn’t use the word commonplace book in his article As We May Think."
- https://bra.in/3vGWLa
- https://bra.in/7v2NRe