Extending Our Perception

Extending Our Perception

The world is rich with information hidden from human senses and not captured by our algorithms. The next era of AI will require Hypersensory Intelligence – AI and novel sensors engineered together to perceive reality in fundamentally new ways, catalysing breakthroughs across disciplines.

What if we could build Hypersensory Intelligence – integrated systems of sensing and AI designed to tackle the world’s most critical challenges?

 

Defined by our Programme Directors (PDs), opportunity spaces are areas we believe are likely to yield breakthroughs.

In Extending Our Perception, we are working to develop Hypersensory Intelligence, where AI is co-designed with novel sensing technologies to directly capture high-dimensional signals from the physical world.

Beliefs

The core beliefs that underpin this opportunity space:

1.

We are under-utilising most sensing modalities when we train AI models → building intelligent systems demands harnessing sensors that extend beyond human intuition to make sense of combinatorial or high-dimensional modalities.

2.

Training on archived data from static sensors will only get us so far → we need dynamic systems that treat perception and cognition as mutually reinforcing and co-evolving components of intelligence.

3.

As we develop Hypersensory Intelligence, symbiotic co-perception will emerge as a new capability class → enabling us to better understand and exploit the evolved biological sensing of humans and other species.

Observations

Some signposts as to why we see this area as important, underserved, and ripe.

Image of ARIA's observations in the Extending Our Perception opportunity space. Tab or scroll down to view the accessible version.

Download as a PDF here, or the accessible version here.

Programme development

As we develop a programme in this opportunity space, we're sharing our current thinking in the form of a programme thesis: "Hypersensory Intelligence: Olfactory Perception".

While AI systems can already see, hear, and generate complex content, they still cannot access a vast and information-rich part of the world: smell. This programme thesis outlines an ambition to build a general-purpose artificial sense of smell, matching or exceeding biological performance on defined tasks, and enabling machines to detect and interpret chemical signals in the real world.

Learn more

Sign up for updates

Stay up-to-date on our opportunity spaces and programmes, be the first to know about our funding calls and get the latest news from ARIA.

Sign up