Protocols for Publishers SHowcase
I'm going to be live skeeting the Protocols for Publishers talks at Newspeak House tonight https://luma.com/2d2s26jg
Welcome to Newspeak House - Protocols For Publishers has a website and a newsletter so look those up.
We're looking to organise events in Perugia, Toronto, LA and maybe Seattle
I'm Ben Werdmuller - lead tech and security for ProPublica before that I built tools based on open protocols
We have publishers putting information out into th eworld to help people make informed democratic decisions
and we have protocol builders who enable people to own their own data and connect it to each other
running tech for a newsroom this last year in the US has been very busy. I have just flown in from a country descending into authoritariaonism. zchildren are being kidnapped, people are being killed
the relationships between publishers and their audience have been intermediated by governments, commerce and advertising
All is not lost. Aroufn the world, mission-driven technologists are working to build an alternative to the prevailing big tech view of the workld
these build togetehr inot what we call the Open Social web - the fediverse, the atmosphere, the indieweb - like the web no-one owns them, no-one can co-opt thtm
publishers could use the sovereign properties of protocols, and potocol peopel could use the pubishers exlpsinign their needs and sharing the results
In NYC last year Chad gathered people who care about both things, and we are here to do the same in the UK
If you're a publisher chat to protocol builders; and vice versa
We're going to start with The Bristol Cable, a community owned newspaper
as a lot of Uk publisher know, after 2012 the trust in mainstrem media was very low, so a groupl of volunteer sin Bristol got togtehr and crowdfunded a newspaper owned by th epeople of Bristol since 2014
We have an open newsroom where the community meets too
referrals are down, discoverability is down, but our membership is growing - Bristol belives in grass-root activism and has a strong political identity
we've had to diversify the kind of journlism we are doing - recntly did a limited series podcast on child incarceration
The Newsmast Foundation is a charity that is 3 years old. Founded after Musk took over Twitter. There were ukrainian communities based on sharing informaton that became unfindable
we have worked through the fediverse and we're now build apps for people
we built thr Bristol Cable app which is both the news app, but also on the fediverse - all mebers of the Cable can read and post on the app, including to the journalists
It allows the mebership to meet each oher, create their own forums and feed inot journalism too.
Building relationships between people is a great way of buildin'g trusst
the Cable news stories are fulled straight from the app membets
you're also connected to the wider social web - you can follow people from the fediverse too.
what it changes is that so far we have interacted with our mrmbers through AGMs, now they can reach us directly
anyone can become a member at £1/ month and join with us
what are you implmening to moderate harassment?
Bristol has a divide between different areas - have you bridged that divide?
we can't say that yet, as we don't know where our embers are but we'd like to do this
Is there a way to bring other publishers in to build out from this?
no yet - we're focused on connecting our members now
we would like to bring more people. Newsmast is working in Quebec to unite multiple publications there to connect community
we developed an App because there was demand fro one - memebrs asked for it
I'm Aendra Rininsland @aendra.com on bsky and I have a talk on how we can make our own algorithms now
I hve worked at the FT, Times, ST Guardian and lots of other publications
I created 4estate,media mastodon instance, which was not succesful
I created the news feed on Bluesky
a decade of algorithmic fuckery: 2015 Google unveils AMP - supposed to make journalism faster, instead cheated pagerank for it
2016: facebook admitted inflated video numbers and he Pivot to Video died. We lost a generation of journalists
2011: Musk takes over twitter and corrupts the results
2023: facebook blocks Canadian news orgs
2024: LLMs replace search with made up results
Why do we still trust Big Tech to do distribution?
Wouldn't it be cool if journalism wasn't blocked by big tech,
Wouldn't it be cool if there was more transarency about where storeis came from
The news feed I made for bsky shows the different news organisation, and trending news delays a bit to find trends
we have about 80,000 Daily averge users of the news feed
there are loads of custom feeds on bsky - Financial time journalist eg and Minenapolis ICE monitors
by making feeds you can build trust and lift the vices of freelancers
if you make a feed that isn't just a single title, there is a broader readership who will pin them
there are lots of ways to make and promote feed - experiment
News orgs can encourage audience to migrate to Bluesky to get more control
bluesky PBC could make custom feeds more visible in the app
remix my news feed https://graze.social/feeds/396
I want to help publishers
democracy dies when the truth is behind a paywall and misinformation is free
anyone can be mass media now - peopel don't subscribe for facts but because you're fighting for truth
paywalls are a retreat from public discourse
Micropaymnets saved the music industry, it can save yours too
legal song downloads didn't save the music industry, streaming subscriptions did
peopel have tried micropayments - wasn't streaming loss of friction, not micropayments?
attempts that will work will be financial systems that we already use inside the browser, and this is more available now
when i was doing ethereum work last year - Indian citizen journalists need this too
people are using payments to subscribe to individual journalists as easily as large institutions - people may trust individuals more
you wre meant to be hearing from E M Lewis-Jong, who is. amuch more charismatic speaker than me so apologies
In early 2021 the first covid vaccines were starting to be approved and wealthy nations were out bidding each other for them - vaccine nationalism
the consequences were that vaccine resistant variants were appearing in pooroer countries and the pandemic continued longer that it needed to
I was working with CoVax in Geneva to try to lift patents and share information more widely - the vaccines came from shared research
right now we are seeing a form of data isolationism in AI development following the same counterproductive pattern
the internet has been used as the source for all AI training, and its putting the internet under strain with all the blocking, licensing and legal standoffs
like the vaccines, this is counterproductive. As long as the public internet is the primary resource for AI, this will continue
The internet is being polluted by AI and it is already unprepresentative. The intreney is less than 1% of the worlds data - most phtos and research data is not there
the abundance is already there - how can we unlock it to make this work?
at The Mozilla Data Collective we have talked to the people who steward these resources and ask them what would make them willing to share?
the biggest barrier was control, not money - who gets to use it uder wha circumstances, and can they cange their minds?
we want genuine governamce over how your data moves through the world - fair value exchange. Could be money, credit, access to tools, or reciprocal access
people want to share their data - The Mozilla Data Collective is in beta, there's a demo but it is still fairly clunky. we're about to spin out of the foundation into a mozilla entity
we have 11,000 accounts as of yesterday - Armenian Q& A discussions, literature in other languages - things people want to shape AI via participation ratehr than extraction
We need to reflect more than English Language content but all of humanity
we're not naive about the challenges - the current Ai favours speed over quality and consent, but we do see a willingness to share on terms chosen by the contributors
We can keep fighting over the contested ground or we can build he infrastructure that makes collaboration possible. If you're holding datasets that woudl make this
is this meant to be an archive? Is labelling/metadata part of this?
more of a marketplace or platform than an archive - labelling is part of this
Mozilla Common Voice, which is labelled voice data in different languages is part of this
how do you handle sensitive or personal data?
through the foundation we can assist groups with doing this - scrubbing PII and manually approving datasets