For the living internet.
Lythaus is a human-first social platform for public-interest conversation, built around clear authorship, accountable participation and feeds you can understand and control.
Join the private betaWe are losing sight of who is speaking and what is real.
The internet has made it easier than ever to publish, connect and participate.
It has also become harder to know what is genuine.
AI-generated content can be produced at enormous scale. Bots imitate participation. Engagement systems reward outrage. Spam competes with original thought. Algorithms decide what reaches us, often without explaining why.
The result is not simply more content.
It is less certainty about who made it, why it reached you and whether the person behind it is willing to stand behind what they said.
Lythaus is being built around a different set of incentives.
Built for humans.
Lythaus is not anti-AI.
AI can be useful. We use it ourselves.
The distinction that matters is between assistance and substitution.
Public conversation carries more meaning when authorship is clear, contribution has accountability and people can distinguish human expression from synthetic generation.
That principle shapes how Lythaus approaches publishing, reputation and discovery.
No public confidence scores. No false precision. Just clearer authorship.
A place for information that matters.
Lythaus is a social platform for people who want more signal and less manipulation in the way they discover and discuss the world.
- Follow people and subjects you care about.
- Share reporting, analysis, experience, questions and perspective.
- Discover conversations across technology, science, business, culture, policy, local affairs, current events and more.
- And have meaningful control over what shapes your feed.
Authorship should mean something.
AI-generated public content isn't permitted on Lythaus.
To support that policy, we are building a layered authenticity system that can consider declarations, provenance, specialist classifiers and behavioural signals.
No single model is treated as an infallible judge of authorship.
Where AI assistance is permitted, it should be disclosed clearly.
Where content is assessed as human-authored, that contribution should be recognisable as such.
And where the system is uncertain, it should be able to say so.
Clear authorship. Not manufactured certainty.
Your feed should not be a black box.
Most social platforms decide what deserves your attention for you.
Lythaus is being designed so that discovery is more understandable and more deliberate.
Your feeds can reflect the people and subjects you choose to follow, while reputation and contribution quality help surface worthwhile discussion.
Discovery should not optimise for
- Not simply what generates the most reactions.
- Not whatever keeps you scrolling longest.
- Not outrage because outrage performs well.
Contribution over attention.
On Lythaus, reputation is intended to reflect how you contribute, not simply how much attention you can attract.
Honest disclosure of permitted AI assistance does not automatically count against you, but it does not earn authorship reputation for work you did not author yourself.
Virality alone is not credibility. Followers alone are not reputation. And volume is not the same as value.
Accountability has to work both ways.
Lythaus will make decisions about authorship, moderation and participation.
Those decisions will sometimes be challenged.
They may sometimes be wrong.
That is why review and appeal are part of the system, with decisions assessed by peers rather than treated as an afterthought.
Authorship and moderation decisions are designed to have understandable reasons, durable records and routes for review.
Community reviewers can contribute to appeals, with additional adjudication for cases that require it.
No person, status or paid tier is intended to be above the rules.
If Lythaus expects accountability from its members, members should be able to expect accountability from Lythaus.
Not everything valuable goes viral.
Lythaus is being built for conversations whose importance cannot always be measured in clicks.
- Original reporting.
- Local knowledge.
- Scientific discussion.
- Analysis.
- First-hand experience.
- Good-faith disagreement.
- Questions worth exploring.
- Ideas that require more than a reaction button.
Public-interest conversation does not have to mean institutional or political conversation. It means information and perspective that can be useful beyond the person posting it.
Credibility should be earned, not purchased.
Lythaus Editorial is intended for journalists, researchers, analysts, academics, independent writers and other contributors doing credible public-interest work.
It is not a paid tier.
It does not grant moderation immunity.
It does not make someone an authority over what is true.
Editorial recognition is earned, reviewable and revocable.
Its purpose is to help people find and recognise credible work more easily, while giving contributors a way to build trust through consistent, transparent and accountable participation.
It can help valuable reporting, research and analysis reach the audiences most likely to benefit from it, without reducing credibility to popularity or payment.
The people behind that work remain accountable to the same platform standards as everyone else.
A different set of incentives.
- We are not trying to maximise how angry you become.
- We are not trying to fill every empty space with content.
- We are not trying to make synthetic participation indistinguishable from human participation.
- And we are not trying to decide what you should believe.
We are building infrastructure for something simpler and increasingly valuable: people speaking for themselves, contributing with accountability and finding information without surrendering all control over how it reaches them.
Lythaus private beta
For the living internet.
Not because everything humans make is good.
Not because technology is the enemy.
Because knowing that another person chose to say something, and is willing to stand behind it, still matters.
Lythaus is being built for that.
Human authorship. Accountable participation. Discovery you can understand.