The concept for Capsule originated from a social media reimagining discussed on Twitter.

The following day, Nadim Kobeissi, a cryptography researcher recognized for creating the now-discontinued, open-source end-to-end encrypted chat application Cryptocat, secured $100,000 in pre-seed funding for his lightweight, mesh-networked microservices idea. Support came from angel investor and former Coinbase CTO Balaji Srinivasan, William J. Pulte, and Wamda Capital.

https://twitter.com/kaepora/status/1348216053030133762

According to Kobeissi, the early-stage startup currently has a post-money valuation of $10 million. He is presently developing a prototype, with the goal of launching a Capsule MVP (Minimum Viable Product) as a web application in March. Following this, he plans to seek a seed funding round, aiming for $1 million to $1.5 million, to expand the team and begin mobile app development.

Currently, Capsule’s presence is limited to a landing page and a pitch deck, which was shared with TechCrunch for review. Kobeissi notes he was surprised by the significant interest in the concept.

“My initial expectation after posting the tweet was that around 60 people would retweet it, potentially leading to a Kickstarter campaign,” he explains. Instead, the tweet “experienced a rapid surge in attention,” resulting in $100,000 in funding “within a single day,” with $50,000 received immediately.

“I don’t identify as a startup founder. My background is in running a business focused on consulting and academic research and development,” he continues. “However, by the end of last Sunday, just eight days ago, I was leading a Delaware corporation valued at $10 million with $100,000 in pre-seed funding – a truly remarkable outcome.”

Capsule represents the newest effort to reshape Internet power dynamics by constructing infrastructure that significantly decentralizes social platforms, aiming to make speech more resistant to corporate censorship and control.

A substantial number of decentralized, P2P (peer-to-peer), and federated protocols and standards already exist, although widespread adoption remains limited. Existing examples include ActivityPub, Diaspora, Mastodon, P2P Matrix, Scuttlebutt, Solid, and Urbit, among others.

Interest in this area has been renewed recently after major platforms like Facebook and Twitter made decisions to restrict U.S. President Donald Trump’s access to their platforms – a demonstration of private power that other political figures have deemed problematic.

Kobeissi shares this perspective, while clarifying that he is not “personally” concerned about Trump’s deplatforming. However, he expresses concern about large private corporations wielding unilateral authority over internet speech, regardless of whether those decisions are made by Twitter’s trust and safety team or Amazon Web Services (which recently suspended the right-wing social network Parler for failing to moderate violent content).

He also references a lawsuit filed in a U.S. court seeking damages and injunctive relief from Apple for allowing Telegram, a messaging platform with over 500 million users, to be available on its iOS App Store – “despite Apple’s awareness that Telegram is used to intimidate, threaten, and coerce individuals,” raising questions about “the likelihood of these efforts gaining traction.”

“That situation is quite concerning,” he suggests.

Capsule intends to mitigate the risk of widespread deplatforming through “easily deployable” P2P microservices, beginning with an upcoming web application.

“Currently, deploying Capsule – I have a prototype that performs minimal functions – involves a single binary file. You obtain that file, deploy it, and run it. It establishes a server, connects to Let’s Encrypt for a certificate, utilizes SQLite for the database, which is a serverless database, and all web server assets are contained within the binary,” he explains, detailing the “compelling technical approach” that quickly secured $100,000 in pre-seed investment.

“There are no additional files – and once it’s running, the folder contains only the Capsule program and a Capsule database file. This self-contained nature makes it easily embeddable and migratable, and achieving this level of simplicity and elegance so rapidly is only possible through this method. For the mesh federation aspect, we’re using HTTPS calls and decentralized database caching.”

Amidst discussions on Twitter regarding how (or if) Kobeissi’s concept differs from existing decentralized protocols, someone shared a link to an XKCD cartoon – which satirizes the tech community’s pursuit of resolving competing standards with a technology that encompasses all use cases (but inevitably increases complexity). Given the number of protocols already offering self-hosted/P2P social media services, it’s reasonable to ask what distinguishes Capsule and why another open decentralized standard is needed.

Kobeissi contends that current options for decentralizing social media are either: (a) not fully P2P (Mastodon is “self-hosted but not decentralized,” according to Capsule’s pitch deck analysis, making its servers “vulnerable to Parler-style AWS takedowns”); or (b) not sufficiently focused on the specific use case of social media (some other decentralized protocols like Matrix aim to support numerous features/apps beyond social media and therefore cannot be as lightweight); or (c) simply too complex for mainstream adoption.

He emphasizes Capsule’s singular focus on social media, similar to Signal’s focus on private messaging, while intending to support both short-form “tweet”-style public posts and long-form Medium-style articles. He is adamant about avoiding “bloat.”

He also references Apple’s “design for usability” philosophy. However, it’s easier to aspire to create something that “just works” than to actually achieve effortless mainstream accessibility. Nevertheless, that is the standard Kobeissi is setting.

“I often envision Glenn Greenwald when I consider my target user,” he says regarding usability, referring to the prominent journalist and Intercept co-founder who recently launched his own newsletter-based platform on Substack. “He’s the person I imagine setting this up. The process would involve him easily setting it up himself or having someone do it for him – I believe Capsule will offer automated deployments as a revenue stream, for a small additional fee, while remaining open source and allowing self-hosting. It won’t be restricted in that regard.

“With Capsule, each content creator has their own website – their own address, such as Capsule.Greenwald.com – and people access content through those addresses, discovering the mesh network through individuals they’re interested in.”

Individual Capsules would be shielded from the risk of platform-level censorship as they would be beyond the reach of takedowns by a single central authority. However, they would still be hosted on the web and therefore potentially subject to removal by their web hosting provider. This means that illegal content on Capsule could still be removed, but there wouldn’t be a single host capable of taking down an entire platform, as happened with Parler and AWS.

“Any takedown would be solely between the Capsule user and their hosting provider,” says Kobeissi. “Capsule users will have a variety of hosting providers to choose from, and each takedown decision will be made by a different entity, potentially with a different judgment. This avoids the centralized control where only Amazon Web Services or Twitter decides who can speak.”

While the web hosting business at a large scale involves a limited number of cloud hosting providers, he argues that this concentration of censorship power diminishes when dealing with numerous decentralized social media instances.

“We have the major hosting providers – like AWS, Azure, and Google Cloud – but beyond them, there are many small hosting providers and businesses. While large businesses often rely on these major providers for their powerful and easily deployable servers, the server resource requirements for running a Capsule instance are generally quite small. In many cases, they are minimal.”

Content would also be more difficult to remove from Capsule because the mesh infrastructure would result in posts being mirrored across the network by the poster’s followers (assuming they have any). Therefore, reposts wouldn’t disappear when the original poster’s account is taken down by their hosting provider. Separate takedown requests would be required to remove each reposted instance, adding complexity to content moderation compared to the unilateral takedowns currently employed by major platforms. The goal is to “protect the community from being silenced,” as Kobeissi puts it.

Trump’s deplatforming appears to have prompted a realization for some that allowing a few corporate giants to own and operate centralized mass communication systems is not conducive to healthy democratic societies, as this centralized control of infrastructure grants them the power to limit speech (and as their content-sorting algorithms determine reach and shape public discourse).

Current social media infrastructure also presents convenient chokepoints for governments to exert censorship.

With growing concerns about the implications of platform power on data flows – and judging by the attention Kobeissi’s tweet received – we may be on the verge of an investor-driven effort to rebuild internet infrastructure to redefine where power (and data) resides.

Notably, Twitter recently renewed its own push for a decentralized social media open standard, Bluesky, indicating a desire to avoid being left behind in any such shift.

“It seems to have really taken off,” Kobeissi adds, referring to his week-old Capsule concept. “I thought I might be the only person who cared. I guess I live in France and am not always aware of what’s happening in the U.S., but many people do care.”

“I’m not a cypherpunk these days, and I don’t advocate for complete anonymity or unaccountability online. If this is abused, we might even encourage hosting providers to adopt guidelines for handling abusive Capsule instances. We want accountability to exist, but we believe it should be organic and decentralized – as originally envisioned with the internet.”