Skip to content
MCAP $2.20T ▼-3.81%
BTC $64,546 ▼-0.96%
ETH $1,814 ▼-0.50%
BNB $592.64 ▼-0.06%
XRP $0.975 ▲+1.99%
SOL $73.87 ▲+2.93%
DOGE $0.0598 ▲+0.65%
ADA $0.155 ▲+1.06%
TRX $0.3074 ▲+0.57%
LINK $9.25 ▲+1.01%
AVAX $7.23 ▲+0.83%
HYPE $69.12 ▲+0.08%
DOT $0.739 ▲+1.00%

The CCS Interview

Swarm director: no open metaverse without decentralized storage

Gregor Zavcer, director at Swarm, the decentralized storage and communication platform that graduated from the Ethereum Foundation, explains how the network works, how it differs from other storage projects, and how its costs and incentives are structured. Recorded in March 2023, the conversation also covers Swarm's push to bring public-good data onto the network and the roadmap Zavcer described at the time.

Interview by Ashton Addison Recorded March 2023 6 min read
Gregor Zavcer Swarm
Gregor ZavcerDirector

Key takeaways

  • Swarm is a decentralized storage and communication platform that functions as a global hard drive where participants contribute storage capacity to a shared network.
  • Unlike other decentralized storage solutions, Swarm users contract with the network rather than individual node operators, providing censorship resistance and data control.
  • Most Web3 applications currently use blockchain for transactions while still relying on centralized Web2 hosting, with less than one percent fully decentralized.
  • Swarm enables diverse layer-two protocols and databases to be built on top, allowing developers to create solutions for personal storage, public datasets, and enterprise use cases.
  • Fair Drive is a decentralized alternative to services like Google Drive and Dropbox that Swarm is developing to demonstrate practical applications of its infrastructure.

From Ethereum's world computer to Swarm

Ashton Addison: How did you get into crypto, and how did that lead you to Swarm?▶ 00:22

Gregor Zavcer: I've been in crypto for ages. The turning point was attending what I think was the first Ethereum meetup in Vienna, before Ethereum launched. Vitalik was presenting the vision of the world computer: the Ethereum blockchain, Whisper messaging and decentralized storage. That made it click for me. Before that I was thinking about decentralized business models, and my environment didn't take it seriously. It was very early. At the same time, as part of my master's studies in cognitive science, I was working on user modeling and was interested in peer-to-peer data solutions, because personal data can be very sensitive and very powerful for those who know what to do with it. Think of Cambridge Analytica, but what worries me more is the things we don't know about. From about 2015 I was completely hooked on blockchain. A year later I met Victor, who we could call Mr. Swarm. Swarm was already an ongoing project within the Ethereum Foundation, and I joined full-time in 2019, when Swarm started its graduation from the Foundation.

Ashton: What is Swarm, and what are its benefits?▶ 03:19

Gregor: Swarm is a decentralized storage and communication platform. Our mission is to empower digital freedom by making data unstoppable. Whoever joins the network contributes storage to it, and that storage becomes part of one global hard drive. It's not like BitTorrent or other decentralized storage where you're just sharing files. Here you become part of one bigger hard drive, which has big implications for what we can build on top of it. Swarm enables a vibrant and diverse layer on top, where protocols and other kinds of databases can be developed, with users in full control of their data. It's also a new use case for enterprises, where all counterparties can access the same data, so they don't need to trust the store of the data.

There's no open metaverse without decentralized storage, and data is just growing in importance.

One infrastructure layer for files and apps

Ashton: Is Swarm more focused on personal file storage or on application storage?▶ 05:09

Gregor: It's all of it, because Swarm is an infrastructure layer. Some developers want to build solutions for personal storage, others want to publish open public datasets, so it gives immense flexibility. People often mention NFTs as a web3 use case, but you can't store the art on a server where somebody can just turn it off. I think the whole domain of decentralized storage is still a bit under the radar. There's no open metaverse without decentralized storage, and data is just growing in importance.

Ashton: Many web3 apps still rely on web2 hosting. How does Swarm's approach to decentralized storage differ from others?▶ 07:13

Gregor: There are different flavors of decentralized storage, built on different models. With some, you have to trust a specific storer node. With Swarm, the person who uploads the data makes a contract with the network, not with an individual storer. That sets the foundation for censorship resistance. It also means node operators aren't liable for the data they store. They don't know what they store, because they hold only parts of data, and those can be parts of many different files. And publishers can be assured they're in full control.

Ashton: Are personal files and applications already being stored on Swarm, and how big is it?▶ 08:45

Gregor: It has a long history, but it's just starting out. Good things take time to build, and I'm very proud of our team. We've been developing tooling and solutions on top of Swarm for quite a while, and they're starting to mature, so now is the time to join. One example is Fair Drive, which is like Dropbox or Google Drive but decentralized. It's also a quick start for developers: take some code, fork it or build your own, connect it to Fair Drive, upload to Swarm, link it to the Ethereum Name Service and you have a dapp running almost out of the box. We haven't tested it, but it's probably a very fast way to deploy. A friend of mine, an expert in decentralization, dubbed this new kind of developer 'weekend warriors.' One developer forks something and deploys it, another comes along and adds a feature, and it can be an ever-evolving cycle. Today, a successful app runs into rising server costs, which force developers to monetize or look for investors, and those investors have their own agenda. With Swarm you just pay for the upload and the storage. I think we can turbocharge not just decentralization but also the open source movement.

Costs, incentives and the IPFS question

Ashton: How are costs determined for decentralized storage, and how do they compare with centralized storage?▶ 12:26

Gregor: I'd say it doesn't really compare. We use blockchain not as a substitute for Google Cloud but for the trust element and the way we deal with money. It's hard to put a price on what censorship resistance or privacy is worth. There are still costs, but the cryptoeconomics are different: the cost of bandwidth and of serving the data is offloaded to users in the form of micro-transactions. For a developer it doesn't matter whether one person or a million download the data or the website, because the network handles it. Swarm also has two types of incentives, for bandwidth and for storage. Nodes compete to be first to serve a chunk of data, so with popular content, nodes cache it and serve it themselves, acting a bit like a content delivery network. The more popular the content, the faster it's served.

Ashton: Is Swarm competing with IPFS?▶ 16:53

Gregor: It depends on the perspective, but I'd argue that no decentralized storage solution is competition to another at this point, because the total market cap and user base of all of them is negligible compared to the big guys. The only way forward for web3 is to collaborate. It would be quite an arrogant attitude to say there's one decentralized storage to rule them all. That would be like saying there's one database for all use cases. The tide lifts all boats: the more popular IPFS is, the better for Swarm, and vice versa. It's a lot easier to explain Swarm to someone who already knows IPFS than to the average person who asks, 'Decentralized what?'

It would be quite an arrogant attitude to say there's one decentralized storage to rule them all.

Enterprises, public data and the 2023 roadmap

Ashton: Does Swarm work with web2 providers so apps can move to web3 without downtime?▶ 19:00

Gregor: Right now we're focused on making our network as performant, robust and secure as possible, and the technology will still take time to mature. The adoption cycle of web3 is also slow. It's unrealistic to expect enterprises to simply switch from one solution to another, because there's a lot of legacy code. I see a transition in which enterprises extend their stack with web3 and decentralized storage. A very good use case in Europe could be GDPR management. There have already been some grants for connecting AWS with Swarm and transferring data easily between the two.

Ashton: Are any big enterprises already working with Swarm?▶ 21:13

Gregor: We're in talks with various organizations. Our goal right now is to support more non-profit organizations, because we believe a lot of good data can be unlocked and shared with the community. At the end of last month we soft-launched data upload sponsorships, and we're preparing an initiative called the Great Data Upload, to first bring humanity's good data onto Swarm and have Swarm act as a public utility. We know community members are bringing OpenStreetMap to Swarm. We're looking at this kind of data first to test the network and to enable options for developers, and that improves the value proposition for enterprises.

Ashton: What are the big things coming for Swarm in 2023?▶ 22:34

Gregor: On layer one, the priority is hardening the network and making it more robust. We're also working on Milestone 4, which brings better support for bigger files. After that we can expect the first protocols to be developed, at least as proofs of concept. On the ecosystem side, which maybe excites me even more, next week we're announcing the graduation of Fair Data Society, an ecosystem initiative we've nourished within Swarm. There's also the Great Data Upload, and more decentralization of the organization itself. I told our team we finally built our car, and now it's time to take it out of the garage and take it for a drive.

This interview has been edited for length and clarity. Recorded March 2023.

Gregor Zavcer
Gregor ZavcerGregor Zavcer is Director at Swarm, an infrastructure provider for decentralized and distributed data storage. He discusses Swarm's positioning relative to existing cloud storage systems and the timeline for Web3 adoption of decentralized storage solutions. Full profile →
Share this interviewXLinkedIn
▶ Watching the interview
CCS Insider · free, twice a week

New founder interviews in your inbox

The week's interviews and the key takeaways, twice a week from CCS Insider.

Readers in 160+ countries. No spam, unsubscribe anytime.