The room that isn't there.

In principle, the internet could have been the answer. In practice, the attention economy ate it. The platforms most of us live on measure success in hours of engagement and ads served. That is their business. It would be strange to expect them to pursue anything else.

What they optimized for was not connection. It was a feeling adjacent to connection that keeps a person scrolling. The difference matters. Ten minutes of real conversation leaves a person more calm than when they started. Ten minutes of feed leaves them more restless. The distinction is not subtle, and it has been studied enough times that no one gets to plead surprise.

There is also a second thing, harder to see from the inside. The products got better at keeping us, and worse at letting us rest. Notifications shaped to produce anxiety. Streaks designed to make absence costly. Feeds recalibrated every few months to favor whatever keeps a user on the surface for a little longer. An entire field of interface design dedicated to the removal of friction from the side of the business, and the addition of it on the side of leaving.

Public libraries exist because literacy is worth too much to leave to bookshops. Public parks exist because fresh air and open ground are a birthright, not a product. Public broadcasting exists because some forms of news cannot survive ad revenue without being warped by it.

Connection belongs on that list. It is too important, and too fragile, to be entirely the business of companies whose business is something else. That is not an argument against private companies doing useful work. It is an argument that a society needs at least one version of this thing that answers to something other than the quarterly numbers.

There are existing examples. Signal runs private messaging as a nonprofit because private messaging is civic infrastructure and should not be owned by an advertiser. Wikipedia is the encyclopedia of record because donors, not sponsors, keep it honest. Mozilla exists because an independent browser is a form of insurance on the open web. Each of them does one important thing in public, with eyes on the books, because the market on its own was going to get it wrong.

We are trying to do that for community and connection.

In the last decade, a small industry has formed around treating loneliness and disconnection as things a person can fix inside their own head. A journal app. A meditation subscription. A mood tracker. A therapist in a chat window. Some of these are good, and they help real people every day. But they are not enough, and in some cases they have let the rest of us off the hook.

Loneliness is not only a private feeling. It is a signal that the surrounding community has thinned out. You can regulate your nervous system until it hums, and you will still need someone to sit across from you at a table. The work of being known is inherently shared. It cannot be installed.

This is also why we are cautious about the clinical frame. Some of the people who need connection most have had their need medicalized into a condition to be managed. We are uninterested in being another app that speaks to people in the language of symptoms. We would rather speak to them as neighbors.

If a Foundation is going to try to do this work in software, the bar is higher than it would be for a company. Here is what that bar looks like in practice. These are not aspirations. They are written into the charter.

No advertising. Ever. Not on the core. Not on a premium tier. Not at the bottom of the feed in small type. The moment a platform carries ads, it answers to the advertiser before it answers to the member.

No selling of individual data. No licensing. No cohort targeting for partners. Aggregated research data only, with explicit opt-in, reviewed by an independent research committee.

Outcomes over hours in the app. We are trying to help a person feel more known. That is not served by making them scroll more. Our internal scorecards, and the ones we publish, measure outcomes, not engagement.

Transparent governance. Audited financials, Form 990, partner lists, executive compensation. Public every year. If it is standard at another serious foundation, it is standard here.

Free at the core, forever. The companion, the communities, the connections, the circle, the wellbeing tools. All of it, free. Membership tiers recognize people who choose to give. They never gate what matters.

A generation is growing up inside these platforms. The harms are legible now in ways they were not in 2012. The research is past the point of debate. And for the first time, the tools to build a credible alternative exist inside the reach of a small, serious team.

There is also a window of public will. Foundations are funding this kind of work. Governments are asking for it. Parents and teachers and family doctors are asking for it, loudly, in every country we have looked at. The thing that has been missing is an organization willing to do the boring, patient, accountable work of building in the open and saying the commitments out loud.

We are going to try to be that organization. None of what we are doing is new in principle. Signal, Wikipedia, Mozilla, and every public library in the country got here first. We are borrowing their posture and applying it to the part of public life that has been left, until now, to the feed.