Artificial intelligence
How intelligent systems learn, reason and adapt.
Research
Intelligence, computing,
and the people using them.
We investigate how intelligent systems can become more efficient, understandable and useful. Our interests connect the models we build, the machines they run on and the ways people work with them.
Explore the writingWhat we investigate
Six connected areas. Shared questions about capability, resources and human control.
Computing architectures
How software and computation can be structured differently.
Efficient & edge intelligence
Smaller, more efficient approaches to useful intelligence.
Multimodal systems
Systems that work across different forms of information.
Human–computer interaction
Ways to make complex technologies understandable and controllable.
Hardware & emerging computing
The relationship between electronics, software and new computing approaches.
From the work
Papers, ideas
and observations.
Research papers, engineering articles and field notes share one archive. These entries are editorial previews while we prepare the first publications.
Technical article Draft
Browse Publications & NotesBringing voice models into one local runtime
The questions behind a shared path from speech to text and back again.
Read the previewResearch paper DraftWhat makes intelligence efficient?
A draft research direction around capability, resources and the machines people already have.
Read the previewField note DraftMaking a circuit easier to understand
An editorial preview about the relationship between a component, a connection and an idea.
Read the preview