jaz-michael king's blog

Fediverse: jaz :twt: :wales_flag: ⁂ (npub1hg6…nwn8)
Blog: https://jaz.co.uk/

Last Notes

2026-05-12 14:21:19 UTC

In my work supporting a broad array of open social web community managers, moderators, administrators, and developers, I regularly hear sweeping generalisations like “the fediverse is anti-this”, “the fediverse doesn’t approve of that”, “the fediverse holds this very specific opinion”.
I’ve always struggled with these statements. How can anyone possibly come to a definitive conclusion about what “the fediverse” believes? As I’ve said before, there isn’t just one fediverse, [there are a million fediverses](https://jaz.co.uk/2025/08/14/there-is-one-fediverse-there-are-a-million-fediverses/ ). You’re only ever subjected to the specific fediverse *you *inhabit, the community you reside on and the people you choose to follow.
If you happen to follow high-signal accounts – people who post prolifically, loudly, and relentlessly – their sheer output completely shapes your reality. Their volume becomes your perception of what the entire network thinks.
For years, we have held up the chronological timeline as our great escape from this kind of distortion. In the open social web, we often point to our lack of an engagement-driven algorithm as a moral high ground. We don’t have a black box sorting our conversations for outrage and engagement, we just have time. First in, first out.
But I’ve long believed this is an incorrect oversimplification. Pure chronological timelines are incredibly easily dominated. They do not naturally create a balanced feed; rather, they inherently privilege whoever has the most time, the most grievance, and the loudest voice.
I’ve struggled to communicate this clearly, but after reading Tobias Rose-Stockwell’s brilliant breakdown at[ The Noisy Room](https://thenoisyroom.com/ ) I’m happy to defer to someone way smarter than me. He outlines a simple metaphor for how commercial social media distorts our reality.
Imagine walking into a pub with a hundred people inside. Ninety-seven of them are having perfectly normal, nuanced conversations. Three of them, however, are screaming at the top of their lungs about politics, about each other, about whatever gets a reaction.
Now, imagine the pub employs a bouncer who gets paid by the minute you spend staring at the spectacle. To keep your attention, the bouncer wires those three screaming people into the pub’s PA system and turns it up to eleven. You walk in, you hear the deafening roar of the three-voiced extremes, and you conclude – logically, based on what you are hearing – that the room is entirely full of unhinged trolls.
In the commercial, centralised web, that bouncer is the algorithm. It amplifies the 3% of users who post severely toxic content because toxicity drives engagement and sells ads.
We look at that and say “thank goodness we fired that bouncer, he’s useless”. We boast that our decentralised, chronological feeds don’t have algorithms manipulating our conversations. But it turns out we don’t need a bouncer to amplify the toxicity via the PA when our timeline does it automatically for us.###
**The Chronological Illusion**When you build a network on a purely chronological feed you replace algorithmic amplification with sheer volume. If those same 3% of vocal, toxic, aggrieved accounts are posting twenty times a day while the 97% of us post once or twice, who dominates your timeline? They do. They don’t need a viral algorithm to amplify them, they just need access to the firehose.
First in, first out simply means the most frequent posters are the most frequently seen/heard.
This creates the exact same distortion that The Noisy Room identifies. We perform an environment scan of our fediverse feeds and conclude that a select few dominant voices reflect the zeitgeist. Maybe we see acrimony, rapid-fire hot takes, relentless indignation, and we believe – falsely – that we are in the minority.
This leads to the same tragic outcomes we see on commercial networks. The quiet majority goes silent. People self-censor. They step away from the keyboard, or they leave the platform entirely, ceding the space to the most extreme voices. The loudest users start to believe they *are* the majority.
Everyone gets each other wrong.
We designed a protocol to save us from algorithms, but we forgot or failed to design for human perception.###
**We need to design for people, not protocols**If we’re going to build better social media, we have to acknowledge that a pure, unfiltered chronological timeline is not a neutral arbiter. It is a megaphone for the relentless. So, how do we enhance the fediverse to fix this?
1. We need client-side enhancements that recognise when a single account is dominating a timeline. If one account posts say ten times in an hour, those posts could collapse into a single stack on my timeline. Let me choose to expand them. Give me back my chronological view of *my community*, not one person’s stream of consciousness.
2. We need to stop treating curation as a dirty word. An algorithm designed by a corporation to maximise ad revenue is harmful, yes. An algorithm, which is just a set of robust filtering tools, designed by *you* to protect *your* peace and balance *your* feed is empowering. We need apps and clients that allow members to easily dial down the volume on highly active accounts without having to fully block or unfollow them, softening the edges without severing relationships.
3. The Noisy Room suggests a “[Community Check](https://github.com/rosestt/communitycheck )“, a representative layer of polling shown below contentious issues to show what the silent majority actually thinks. While this would be complex to implement across a decentralised network, we can build tools that gauge consensus without relying on the loudest voices. We need to find ways to measure and display community sentiment that isn’t just counting the number of angry, rapid-fire replies. Community managers should be able to [add community notes](https://github.com/mastodon/mastodon/issues/27302 ) to posts, or reply with them in a fashion that pins them to the OP. Emelia and the W3 SWCG Trust and Safety taskforce [has some thoughts on this](https://github.com/swicg/activitypub-trust-and-safety/issues/4?ref=writings.thisismissem.social ).###
**You’re not in the minority**Most people want their own space, shaped by their needs and their values. They want to connect, share, and learn without being shouted at. The social web is full of these people. I’m one of them. We are the 97% having a normal conversation in the pub while the 3% scream near the bar.
Our goal shouldn’t just be preserving a chronological feed at all costs. Our goal should be building better relationships. We need to ensure that when a new member joins any one of our million fediverses, they see the whole room, not just the people shouting the loudest.
Let’s find our common ground, let’s build tools that reflect the reality of our communities, and let’s give the quiet majority their voice.
https://i0.wp.com/jaz.co.uk/wp-content/uploads/2026/05/image.png?fit=635%2C271&ssl=1

2026-03-13 15:55:49 UTC

We’re at a bit of a strange moment in the history of the Internet. Long-standing assumptions about platform freedoms and responsibilities are being tested by a wave of regulation from countries far and wide, including extranational jurisdiction where all of a sudden people are finding themselves subject to laws of countries they’ve never been to.
The age of self-regulation is ending. In 2021 there were three OECD safety regulators. Today, a mere five years later, there are 28.
The two-year-old [Global Online Safety Regulators Network](https://www.esafety.gov.au/about-us/consultation-cooperation/international-engagement/the-global-online-safety-regulators-network ) brings together independent regulators to cooperate across jurisdictions, which may lead to [alignment across nations and regulations](https://www.esafety.gov.au/sites/default/files/2025-11/GOSRN-Regulation-Index-2025.pdf?v=1773412565992 ), from Fiji to France.
For social media alone, [three countries have implemented age restrictions](https://social-media-age-tracker.onrender.com/#status=implemented,passed ), five have passed laws ready to implement, and 30 more discussing or under consideration at a national level. For many of us, it’s jarring and unsettling, and this is especially true if you’ve never worked in a regulated industry before.
Internet regulation is generally converging around three main themes:<li>Platform integrity: the macro responsibility for elections and social discourse.</li><li>Individual safety: the proactive duty of care for children and vulnerable people.</li><li>Personal agency: mandating data portability and banning manipulative dark patterns.</li>
These new burdens feel sudden and invasive if you’ve spent your career inside the Web’s long-running freedom experiment. For me, however, they follow a pattern I’ve seen before. I think there are some lessons we can learn from – and be better prepared – by looking beyond digital media.##
The Myth of ExceptionalismI started working on the Web in the mid-nineties. It was liberal in the most literal sense, few gatekeepers, no meaningful regulation, and a shared belief that this new thing would disintermediate everything. Media, power, and information were all up for grabs.
“Under construction” signs abounded. Dancing dog gifs ran across Web pages. The Clinton White House [specifically told everyone to go nuts](https://clintonwhitehouse4.archives.gov/WH/New/Commerce/read.html ), that no amount of regulation could possibly keep up with how fast the Net was innovating. Phrases like “The private sector should lead” and “Governments should avoid undue restrictions” stimulated a self-regulating industry that is now bristling at unwanted government interference.
We built first and worried later. The default assumption was freedom; responsibility was personal rather than systemic. We created a mythos that the Internet was exceptional, as if it were somehow immune to the rules of the offline world.
Then, in 2001, almost by accident, I moved into healthcare.##
When Regulation Provokes OutrageHealthcare is one of the most heavily regulated environments on the planet, and for good reason – lives are at stake. My transition into this world coincided with a wave of mandates designed to break open the black box of medical outcomes. From 1995 to 2000, a range of reforms hit healthcare, built on the back of seminal writings like “[To Err is Human](https://pubmed.ncbi.nlm.nih.gov/25077248/ )” and “[Crossing the Quality Chasm](https://pubmed.ncbi.nlm.nih.gov/25057539/ )“, leading to the creation of state and national report cards on nursing homes, hospitals, physicians and more.
I became one of the people responsible for turning those legal requirements into reality. I was building the databases and public dashboards that the laws now required to make healthcare quality and cost visible to the public. I built, provided data for, consulted or advised on just about every initiative in the US in that decade. I designed and deployed [a national data infrastructure](https://pellucid.atlassian.net/wiki/spaces/PEL/overview?homepageId=8028166 ) for codifying performance, cost and quality data over time. I made a [lot](https://healthcarereportcard.illinois.gov/search ) [of](https://whynotthebest.org/ ) [report](https://profiles.health.ny.gov/nursing_home/index#5.79/42.868/-76.809 ) cards. This made quite a few people very angry.
Clinicians and administrators confronted me directly. They claimed the data was unfair or lacked context. What struck me was that the anger was rarely about the underlying disparities in care. Instead, it was about the act of publication itself, which was now compelled by law. Being made transparent and accountable to the public felt like an attack on their professional identity.
I see that same identity shock today from technologists as they realise their closed systems and unfettered independence are being opened and constrained by similar regulatory tools.##
From Hero Culture to ComplianceAs healthcare matured, we moved from transparency to interoperability. This arrived with [HIPAA](https://www.hhs.gov/hipaa/index.html ) in the US. At the time, it was loathed. Technologists complained it was vague, hard to implement, and poorly aligned with how software worked.
Most people assume the P in HIPAA stands for privacy; it actually stands for portability. Regulation did not stop data sharing. It professionalised it, it shifted the focus from institutional protection to patient-centredness.
We saw the same arc in cybersecurity, as privacy and security became more entwined. It began as a bit of a dark art, which was the era of the hero admin. But as harms and threats accumulated, an industry formed, best practices became auditable compliance. [SOC 2](https://www.aicpa-cima.com/topic/audit-assurance/audit-and-assurance-greater-than-soc-2 ), [ISO 27001](https://www.iso.org/standard/27001 ), [FISMA](https://www.congress.gov/bill/113th-congress/senate-bill/2521 ) became common parlance. Risk assessments and incident reporting stopped being optional. Cybersecurity and the regulation that requires and enforces it did not kill innovation, it gave us a shared language for risk, and has had demonstrable benefits to society. Data breach disclosure laws, for example, [resulted in a significant drop](https://onlinelibrary.wiley.com/doi/abs/10.1002/pam.20567 ) in identity theft incidents. The ID Theft Center reports a [major decline in breach notifications](https://www.idtheftcenter.org/wp-content/uploads/2026/01/2025-ITRC-Annual-Data-Breach-Report.pdf ), thanks in the main to data privacy regulations (and the fines that come with ignoring them).##
The Internet is Catching UpThe Internet is now at this exact same tipping point. What cybersecurity was to data integrity, modern regulation is to platform safety and rights. The idea of a permanently unregulated Internet is simply not credible anymore.
This is not because regulators hate our freedoms. It’s because the Internet is a fundamental and vital utility that now shapes childhood, mental health, and physical safety. When systems reach that level of impact, they get regulated. Period. Whether it’s [Australia’s social media ban for minors](https://www.unicef.org.au/unicef-youth/staying-safe-online/social-media-ban-explainer ), the [EU’s Digital Safety Act](https://digital-strategy.ec.europa.eu/en/policies/digital-services-act ), or the[ UK’s Online Safety Act](https://www.gov.uk/government/publications/online-safety-act-explainer/online-safety-act-explainer ), it’s not even a case of “censorial Europe” versus a “free” America. [Half of all US states now require age verification](https://www.404media.co/missouri-age-verification-law-porn-id-check-vpns/ ) for adult content. This is political reality catching up everywhere at once.##
Fediverse OperatorsIf you run a small server, perhaps a local community hub or a niche interest instance, this can feel incredibly intimidating. Regulation sounds like something designed for platforms with a thousand lawyers and a big shiny office in California.
However, most of these laws, and the regulators that enforce them, follow a [principle of proportionality](https://www.bihr.org.uk/get-informed/legislation-explainers/what-is-proportionality ). A regulator’s actions must be lawful, legitimate, and proportional. Lawful means there must be a law allowing official to take action, legitimate means there is a good reason to act, and proportionate means there is no other way to seek the desired outcome, it is the least restrictive option available to the regulator.
You are not being measured against the same yardstick as a multi-billion-dollar platform. The goal is good faith, not perfection. For small services, compliance usually looks like a few clear signals:<li>Be reachable: have a clear way for authorities and users to contact you on your About page.</li><li>Be clear: document your policies and how someone can appeal a decision.</li><li>Be informed: use the resources the community is already building.</li>
If a regulator ever comes knocking, they won’t ask why you missed one bad post. They’ll ask “Do you have a process for when things go wrong?” Make sure you have one. And as a start, copy someone else’s. Find a service you like and look at their documentation. Beg, borrow, or steal liberally. [Here’s everything I have on toot.wales](https://croeso.toot.wales/en/policies/ ) for example, help yourself.
The early Web was a gift, but freedom without accountability does not scale. There is one Fediverse, there are a million Fediverses, and there are a thousand regulations each of them has to be ever so gently aware of. Learning to live with a regulated Internet is not a failure of imagination. It’s the next phase of our shared responsibility.##
Further Reading<li><a href="https://www.google.com/search?q=https://about.iftas.org/dsa/" target="_blank" rel="noreferrer noopener">IFTAS DSA Guide</a> – Guidance specifically for small providers. It breaks the Act down into manageable, actionable steps for Fediverse admins.</li><li><a href="https://decoded.legal/blog/" target="_blank" rel="noreferrer noopener">Decoded Legal</a> – This is a brilliant resource for understanding how English law views decentralised services. Their analysis of the UK Online Safety Act is essential reading for anyone running a UK-based instance (and can be insightful for others too).</li><li><a href="https://fediverse-governance.github.io/images/quick-start.pdf" target="_blank" rel="noreferrer noopener">The Fediverse Governance Guide</a> – A community-led effort to document best practices for server moderation, transparency reports, and user appeals.</li><li><a href="https://about.iftas.org/wiki_cats/legal-regulatory/" target="_blank" rel="noreferrer noopener">The IFTAS Community Library</a> – A growing repository of tools and documentation templates. You can find examples of clear, legally-aware content you can adapt for your own community.</li><li><a href="https://www.google.com/search?q=https://www.eff.org/issues/bloggers/legal/osp" target="_blank" rel="noreferrer noopener">Electronic Frontier Foundation (EFF) – Legal Guide for Online Service Providers</a> – While US-centric, this provides a great foundational understanding of the protections that shaped the Internet and how they are evolving today.</li>
https://i0.wp.com/jaz.co.uk/wp-content/uploads/2026/03/image.png?fit=816%2C491&ssl=1

2026-01-13 15:22:37 UTC

Recently, I was sitting with my sister-in-law talking about social media. She has that nagging, familiar concern about corporate social: the feeling that we are all just products in someone else’s machine. I pulled out my phone, showed her my toot.wales feed, and started explaining how she could “save the web site to her home screen”.
She listened politely, then said the thing I believe most people genuinely feel:
**“It’s lovely! If only there was an app…”**
She does not want a “protocol”. She does not want to hear about “the Fediverse”. She just wants *that *feed on *her* phone without a homework assignment.
When I started toot.wales, I knew this instinctively. In the early days, I even paid a developer to fork a couple of open-source apps and hard-code them to our server. It was expensive and eventually I simply couldn’t sustain it, but I never lost sight of that goal.
Now, working with our brilliant friends at the [Newsmast Foundation](https://www.newsmastfoundation.org/ ), we have finally delivered it: a simple, opinionated, “people-first” social app.###
One community. One door.The biggest design decision we made was the simplest: the app is hard-locked to toot.wales.
There is no “pick a server” screen or “find your instance”. There is no friction. This is not “a Mastodon app”; it is the Tŵt app, purpose-built for our community.
I’ve spent years building tools like *[StartHereSocial](https://jaz.co.uk/projects/startheresocial/ )* to help people navigate the “maze” of the Fediverse. However, for toot.wales, I realised the answer should not be a better map. It should be an inviting, open front door. If you want to join us, you should be able to step right in.###
Croeso: A warmer welcomeAlongside the app, I built a promotional web site at [croeso.toot.wales](https://croeso.toot.wales ). *Croeso* is Welsh for “welcome”, and that is exactly what it provides. It strips away the jargon. There is no talk of federation or decentralisation. Instead, it offers clear language about why this is “Better Social”.
It is joyful and unthreatening on purpose. We are not here to preach at you to quit X or Facebook, and we are not shaming anyone. We are simply inviting you to try something different. If you like the vibe, stay. If not, that is fine too. People stay because it feels good, not because they have been talked into a boycott.###
A Fediverse of communitiesSome people see the Fediverse as one giant global town square. I see it differently. For me, it is a constellation of communities.
When people tell me “mastodon.social is too big”, my gut feeling is no, *your* server is too small. Communities only grow when we intentionally nurture them. It is not up to a company in Germany to grow Tŵt. That’s on us, it’s our problem.
I want everyone in Wales (and the world!) to have a social option that is not drowned in surveillance and ad-tech. To get there, we have to design for clarity rather than complexity.###
What’s inside the Tŵt app?
We have made some very deliberate choices to keep the app “human-shaped”:<li>Private Mentions have their own space. You cannot create a private mention in the main composer and you will not see them in your timeline. We have put them in a dedicated “Conversations” tab to reduce mistakes and keep privacy intentional.</li><li>Three simple tabs. You get Home (your personal feed), Tŵt (the “local” feed plus replies), and Explore (your lists and hashtags, plus curated feeds for news, sport, culture and more). This creates a natural flow from your own interests to the wider community.</li><li>Search is just search. There are no “trending” nudges or unsolicited content. If you are looking for something, we help you find it. We don’t try to distract you along the way. Trending is it’s own Channel.</li><li>True Bilingualism. English and Cymraeg sit side-by-side as equal choices. Digital spaces should reinforce our identity, not flatten it.</li>###
Celebrating the pioneersThe Tŵt ecosystem is already thriving because Welsh media organisations took a leap of faith into the open web. Groups such as Nation.Cymru, Golwg360, Swansea Bay News, and Wrexham.com did not wait to see which way the wind was blowing. They just started building, and are long-time social web citizens.
Whether it is weather updates from North Wales Storms or local sports teams, these partners make the app feel “live” from the moment you log in. The app offers two pathways to consume this content, via Channels you can drop in and out of, or Starter Packs you can follow to bring them into your feed.###
Why this mattersFor those already comfortable with the Fediverse, using a web interface is fine. But for everyone else, “just install the web site” is not meaningful advice.
People want “the app”. They want to tap an icon and feel at home, rather than feeling like they are doing a technical chore.
If you run a community and like our approach, please feel free to use it. Whether it is the onboarding flow, the “Better Social” messaging, or the focus on language equity, you are welcome to use my work as a template. If we want a healthier internet, we need more communities feeling empowered to build their own front doors.
The Tŵt app is our doorway. It is built for Wales and the Welsh, at home and abroad. It is simple, it is ours, and I cannot wait for you to try it. We plan on launching on Dydd Dewi Sant, a special day for Cymru, in the meantime you can get a preview at https://croeso.toot.wales/en/less-clutter-more-cwtch/
https://i0.wp.com/jaz.co.uk/wp-content/uploads/2026/01/Twt-Cymru-Toot-Wales-App-Store-Summary.png?fit=1024%2C500&ssl=1