Can social media start over? Bluesky’s CEO and COO deliver their case at TechCrunch Disrupt 2026

Bluesky is not trying to build a better Twitter. It is trying to make the concept of "a Twitter" obsolete. That distinction matters. At TechCrunch Disrupt 2026, CEO Toni Schneider and COO Rose Wang will argue that social media's next act belongs to an open protocol — the AT Protocol — not another walled garden. If they're right, the map of the social web redraws itself. If they're wrong, Bluesky becomes another Ghost, another Mastodon, another noble experiment that settled for niche.

Schneider took the permanent CEO role in July after four months as interim chief. The transition was deliberate. Founding CEO Jay Graber moved to chief innovation officer, explicitly ceding the operator's chair. His line: Bluesky needed "a seasoned operator focused on scaling and execution." Translation: the protocol works. Now prove the business can.

Wang has been inside since late 2021. She ran customer experience at Forethought AI before that. Her view is sharper than the standard founder rhetoric. At SXSW London she said the quiet part aloud: Facebook and X are basically AI companies now. Their feeds are synthetic. Their optimization targets are engagement metrics that reward outrage and retention, not connection. Turning that ship requires rewiring their revenue engine. They won't.

That is Bluesky's opening. Not feature parity. Not a cleaner timeline. The opening is structural. The AT Protocol separates identity, data, and application layer. A user's social graph — followers, posts, likes — lives at the protocol level. Any client can read it. Any developer can build a new interface on top. Bluesky the app is merely the first reference client. The network effect compounds at the protocol layer, not the app layer. That is the theory. The practice is messier.

Early adopters know the friction. Onboarding still asks non-technical users to understand "PDS" and "DID" and "lexicons." The firehose of content is thinner than X's. The culture is earnest, sometimes performatively so. But the architecture holds. When a third-party client like Skyfeed or Graysky ships a feature Bluesky's own team hasn't built, the protocol argument gains evidence. When a researcher pulls the full public graph for a study on misinformation without negotiating API access, the argument gains teeth.

Wang's human-connection thesis is the commercial wedge. She bets that a cohort of users — creators, professionals, communities burned by algorithmic whiplash — will pay for a feed they control. Not with money necessarily. With attention. With loyalty. With the willingness to rebuild their graph from scratch. That rebuild cost is the moat. Bluesky must lower it. Portable identity helps. So does the "starter pack" feature that imports a curated follow list in one click. But the graph is still social capital. People do not abandon capital lightly.

Schneider's decade timeline is honest. Protocol adoption curves are measured in years, not quarters. Email took two decades to become universal. The web took one. ActivityPub — the protocol behind Mastodon — has run since 2018 and still rounds to zero in mainstream consciousness. AT Protocol is younger. Its governance model, a consortium of independent operators, is untested at scale. Bluesky PBC (Public Benefit Corporation) holds the trademark and the reference app. The protocol specification lives in a separate foundation. That split is designed to survive Bluesky the company. Whether it survives Bluesky the company's investors is an open question.

The incumbents are not blind. Meta's Threads adopted ActivityPub in beta. X open-sourced its recommendation algorithm, then quietly stopped updating the repo. Both companies have the distribution to crush a protocol transition if they committed. They haven't. Their business models depend on owning the feed. Owning the feed means owning the data. Owning the data means closing the protocol. The conflict is structural.

Bluesky's revenue model remains vague. No ads. No promoted posts. Schneider has hinted at "services on top of the protocol" — custom domains, verification, enterprise tooling, perhaps hosted PDS instances. That is plausible. It is also the business model of every open-source infrastructure company that eventually sells to a cloud vendor. The difference: the asset here is a social graph, not a codebase. The lock-in is human.

Wang's candor about AI-filled feeds resonates because it matches lived experience. The "following" tab on X is buried. Instagram's chronological feed is a buried setting. TikTok never had one. The algorithm is the product. Bluesky's reverse-chronological default is a political act disguised as a UX choice. It says: you decide what you see. The protocol says: you decide who decides.

At Disrupt, Schneider and Wang will face a room of founders who have watched a dozen "Twitter killers" launch and fade. The skepticism is earned. But the protocol bet is different in kind, not degree. It does not require Bluesky to win. It requires the protocol to become the substrate. If Skyfeed wins, or a client built by a newsroom, or a DAO-governed community client, the protocol wins. Bluesky the app can become the Netscape of the social web — first, formative, eventually irrelevant — and the thesis still pays out.

That is the case they must make. Not "we are the next big platform." Not "join our waitlist." The case: the architecture of the social web is up for grabs. The protocol is the only lever that moves the incumbents. Everything else is a feature request.

Schneider's decade clock started ticking in July. Wang has been watching the user sentiment shift since 2021. The Disrupt stage is their first major joint prosecution of the argument. The evidence will not be a demo. It will be the developers building on AT Protocol without permission. The communities migrating their graph because they can. The researchers studying the network because it is legible. The protocol either compounds or it stalls. Bluesky's leaders are betting their careers on compound interest. The room at Moscone West should bet on the structure, not the company.