# Master's thesis at DISCO (2022–2023)

> Back to hardcore distributed systems for the master's thesis: the "world computer" vision, finding the Internet Computer, and developing a faster consensus protocol at ETH's Distributed Computing Group that became Banyan — Best Paper at Middleware 2024.

- Canonical URL: [https://massimo.context-use.com/p/about/chapters/2020-2024_zurich/2022-2023_master-thesis-at-disco](https://massimo.context-use.com/p/about/chapters/2020-2024_zurich/2022-2023_master-thesis-at-disco)
- Site index: [https://massimo.context-use.com/llms.txt](https://massimo.context-use.com/llms.txt)
- Last edited: 2026-07-26T11:07:26.637Z

After completing my semester project at the University of St. Gallen and my internship at EVRYTHNG — both focused on [the Web of Things and ubiquitous computing](https://massimo.context-use.com/p/about/chapters/2020-2024_zurich/2021-2022_web-of-things/story) — I decided that for my master's thesis I wanted to go back to hardcore distributed systems.

### The world computer

By then I was a big believer in [digital sovereignty](https://massimo.context-use.com/p/about/chapters/2020-2024_zurich/2020-2021_getting-into-distributed-systems), and I thought it should extend way beyond digital money (which Bitcoin had enabled). Systems like Ethereum had been built with the vision of creating a "world computer": a network of computers managing a blockchain that could execute any computation (i.e. run applications), not just bookkeeping a ledger of digital currencies.

A "world computer" managing a blockchain would provide properties found in none of the individual computers — once again an instance of the emergent behaviour I had fallen in love with during [the study of complexity theory](https://massimo.context-use.com/p/about/chapters/2017-2020_politecnico/story). One of the most fascinating properties of a "world computer" would be the creation of networks (as in social networks) where the value generated by the network accrues to the edges (the many people participating, contributing, and creating the value) as opposed to the center (the few people who bootstrapped the network). The book [Read Write Own](https://readwriteown.com/) by Chris Dixon best explains why such a property would ensure a better future for the web — and for all of us.

### Finding the Internet Computer

Anyway, the more I learned about Ethereum, the more I realized that its architecture would not be able to accommodate a "world computer" capable of hosting real networks and other applications. That's when I started looking for alternatives.

Eventually, thanks to a VC, I found out about the [Internet Computer](https://internetcomputer.org/), a blockchain created by [DFINITY](https://dfinity.org/), a Swiss foundation headquartered in Zurich.

What got me interested in the Internet Computer was the fact that they had decided to sacrifice a bit of decentralization (the Internet Computer is technically not a permissionless blockchain) in order to get massive gains in efficiency — thus getting closer to a practical "world computer". Reading their papers on [the consensus algorithm](https://eprint.iacr.org/2021/632.pdf) and [the system as a whole](https://internetcomputer.org/whitepapers/The%20Internet%20Computer%20for%20Geeks.pdf), it became clear to me that they were best positioned to realize the vision of a "world computer". This is what I wanted to research for my master's thesis.

### Banyan

Luckily enough, a couple of PhD students at ETH's [Distributed Computing Group (DISCO)](https://disco.ethz.ch/) had recently published a paper called [Consensus on Demand](https://tik-db.ee.ethz.ch/file/73b619fed17b1376e6d14dee8732345f/), which reduced the latency of permissioned consensus protocols, and they were looking into the Internet Computer as a potential candidate to apply their algorithm to. That's how I ended up developing [a faster consensus protocol for the Internet Computer](https://arxiv.org/pdf/2312.05869v1), which eventually became [Banyan](https://dl.acm.org/doi/abs/10.1145/3652892.3700788) and won the [Best Paper Award at Middleware 2024](https://ee.ethz.ch/news-and-events/d-itet-news-channel/2024/12/best-paper-award-for-banyan.html) *[link to the dedicated project page to come]*.