R01
Artificial Intelligence
Reasoning, planning and learning systems, and the question of where their judgement should stop.
Kosmia Research
Kosmia Labs explores emerging technologies before there is a product to attach them to. Some of it becomes engineering. Some of it stays a question. Both are worth the time.
R01
Reasoning, planning and learning systems, and the question of where their judgement should stop.
R02
Turning raw sensor data into a stable understanding of a scene, under conditions nobody planned for.
R03
The loop from perception to planning to control, and what breaks when the physical world disagrees.
R04
Long-running systems that make decisions without a person watching, and stay predictable while doing it.
R05
Interfaces for systems that think back. How intent is expressed, and how machine reasoning is made legible.
R06
Work that is repetitive but not simple, and the judgement required to hand it over safely.
R07
Software that holds a model of physical space, and interfaces that live inside it rather than on a screen.
R08
The next area is not chosen yet. It usually arrives as someone's side experiment.
/ Notes
We plan to publish what we find, including the experiments that fail. Nothing is written yet, so there is nothing to link to.