WikiHouse

WikiHouse is a digitally-manufactured building system that makes it simple for anyone to build beautiful, zero-carbon homes.

Translating Neuralese

Several approaches have recently been proposed for learning decentralized deep multiagent policies that coordinate via a differentiable communication channel. While these policies are effective for many tasks, interpretation of their induced communication strategies has remained a challenge. Here we propose to interpret agents' messages by translating them. Unlike in typical machine translation problems, we have no parallel data to learn from. Instead we develop a translation model based on the insight that agent messages and natural language strings mean the same thing if they induce the same belief about the world in a listener. We present theoretical guarantees and empirical evidence that our approach preserves both the semantics and pragmatics of messages by ensuring that players communicating through a translation layer do not suffer a substantial loss in reward relative to players with a common language.

Learning to Communicate with Deep Multi-Agent Reinforcement Learning

We consider the problem of multiple agents sensing and acting in environments with the goal of maximising their shared utility. In these environments, agents must learn communication protocols in order to share information that is needed to solve the tasks. By embracing deep neural networks, we are able to demonstrate end-to-end learning of protocols in complex environments inspired by communication riddles and multi-agent computer vision problems with partial observability. We propose two approaches for learning in these domains: Reinforced Inter-Agent Learning (RIAL) and Differentiable Inter-Agent Learning (DIAL). The former uses deep Q-learning, while the latter exploits the fact that, during learning, agents can backpropagate error derivatives through (noisy) communication channels. Hence, this approach uses centralised learning but decentralised execution. Our experiments introduce new environments for studying the learning of communication protocols and present a set of engineering innovations that are essential for success in these domains.

Emergence of Grounded Compositional Language in Multi-Agent Populations

By capturing statistical patterns in large corpora, machine learning has enabled significant advances in natural language processing, including in machine translation, question answering, and sentiment analysis. However, for agents to intelligently interact with humans, simply capturing the statistical patterns is insufficient. In this paper we investigate if, and how, grounded compositional language can emerge as a means to achieve goals in multi-agent populations. Towards this end, we propose a multi-agent learning environment and learning methods that bring about emergence of a basic compositional language. This language is represented as streams of abstract discrete symbols uttered by agents over time, but nonetheless has a coherent structure that possesses a defined vocabulary and syntax. We also observe emergence of non-verbal communication such as pointing and guiding when language communication is unavailable.

Ambient Agoras: Dynamic Information Clouds in a Hybrid World

One project in the new proactive initiative

The information percolator: ambient information display in a decorative object

As part of the Digital Library's transition to Open Access, new features for researchers are available in the Premium Edition. Click here to learn more.

Locating Family Values: A Field Trial of the Whereabouts Clock: (Nominated for the Best Paper Award)

Static! Designing for energy Awareness

From the boost of the first cup of coffee in the morning to the glow of a nightlight that keeps a child company after dark, much of our contemporary lifestyles is powered by electricity. Today, however, the economic and environmental costs of energy require us to think again. While recognizing the complex social functions and cultural forms of electricity in everyday life, the current challenge for design is to change practices and patterns of (over)consumption. Static! explores design as a basis for increasing awareness of energy. Familiar furnishings and products have been reinterpreted to materialize electricity, to make it more visible and tangible for people. The resulting series of conceptual design examples express the poetics – and politics – of everyday electricity consumption. A design research project at the Interactive Institute funded by the Swedish Energy Agency, Static! has proved to be a pioneer in opening a current and critical research area at the intersection of design, energy and information technology. This book presents the Static! design examples and perspectives on issues in (sustainable) design today. Grounded in passion and humor, as well as rigor and research, the book asks designers and consumers – that is, all of us – to rethink the form, and future, of electricity in the world around us. Editor: Ramia Mazé. Authors: Christina Öhman, Sara Backlund, Johan Redström, Ramia Mazé, Sara Ilstedt Hjelm, Sara Routarinne.

FOSDEM 2025 - Goblins: The framework for your next project!

Building peer-to-peer decentralised applications remains difficult and error-prone. Most attempts at this either abandon collaborative features entirely or fall back on centralised architectures. The Spritely Institute is working on this challenge by creating (among other things) Goblins, a Guile framework that makes secure, fault-tolerant peer-to-peer applications accessible to developers. These tools are especially valuable for developers building secure collaborative applications that aim to foster healthy online communities. This talk walks you through Goblins’ most powerful features, including the actor model, object capability security, networking, time travel debugging, and persistence.

FOSDEM 2026 - How to level up the fediverse

The fediverse isn't done! Or, at least, that's the opinions of ActivityPub co-authors Christine Lemmer-Webber and Jessica Tallon! Discover how we could improve the fediverse with ActivityPub-compatible improvements to add more robust and secure communication/cooperation patterns, decentralized storage and identity, etc! Plus: how Spritely is starting to apply its technology towards exactly this goal, bringing Spritely's next-generation internet tech to the fediverse!

Plan B: an operating system for ubiquitous computing environments

The conventional approach for building pervasive environments relies on middleware to integrate different systems. Instead, we have built a system that can deal with these environments by exporting system resources through distributed virtual file systems. This requires no middleware, simplifies interoperation, and permits applying general purpose tools to any system resource. A constraint-based file system import mechanism allows the system to adapt to changes in the environment and permits users to customize the environment and tailor adaptation according to their needs. The system has been in use for over a year to carry out our daily work and is underlying the smart space that we built for our department

The Plan B OS for ubiquitous computing. Voice control, security, and terminals as case studies

MICROCOSM: An Open Hypermedia Environment for Information Integration.

This report examines open hypermedia systems, and argues that such systems provide users with richer and more diverse ways to access and integrate information from large and dynamic data sets in a distributed, heterogeneous environment. In particular, the enhanced Microcosm model for open hypermedia is examined, and the ways in which it provides such an environment are discussed. The paper continues by investigating the advantages and the short comings of this model and identifies areas in which further work must be completed before such systems can become widely adopted, in particular the granularity of link anchors, editing, and version control. Possible solutions to these problems are presented and discussed.

Vision and Reality of Hypertext and GUIs: 2.1.13 Microcosm @mprove

Microcosm (cf. 2.1.13) is a research project at the University of Southampton. Like Intermedia five years before, Microcosm is designed around the idea that material of diverse sources and formats should be tied together into one hypertext. To achieve this goal links have to be separated from the documents. They are stored and maintained in special link databases. In contrast to Intermedia, Microcosm plays the active role in gathering the necessary data out of the documents. For each topic it stores data on which documents contain information about it and also where inside the documents the information is located. This elaborated conception of links can be seen more as a kind of Memex’ trails than simple one-to-one links. Given this structure dynamic linking becomes possible. This means that links can be associated with generic text strings. Wherever this string shows up a link is placed automatically. This has the effect that documents that are imported for the first time into Microcosm can have links immediately.

Do we actually need to write research papers at all?

If we are going to lean on the creativity and the cross-domain thinking of human researchers, then surely that should be where the human time goes.