Frontier AI · Nonhuman intelligence

Life is
transmitting.
We’re learning
how to listen.

We build AI that interprets signals from non-human life—beginning where they can be measured, tested, and made useful.

Signal field 001
Biological time-series / live rendering
LISTENING AT THE EDGE OF INTELLIGENCE
37° 48′ 12.4″ N
A new instrument for life
01 / The thesis

Intelligence did not begin with us. Communication does not end with language.

Across the living world, information moves through sound, gesture, chemistry, electrical activity, and behavior. We are building the models and instruments to make these signals legible—without pretending we already know what they mean.

Founded by former DeepMind researchers, our lab brings frontier machine learning into direct conversation with biology, field science, and careful experimentation.

01 / MEASURE
Start with the signal.Capture rich, synchronized observations in the environments where life actually happens.
02 / INTERPRET
Model with context.Connect sound, motion, physiology, and environment—not isolated clips stripped of meaning.
03 / VALIDATE
Earn every conclusion.Test interpretations against behavior, repeatable experiments, and domain experts.
02 / The research engine

From raw signal
to tested meaning.

A shared intelligence stack for very different forms of life. Species-specific data and science. Reusable tools for listening, learning, and validation.

01 — SENSE

Observe

Build non-invasive sensor systems for sound, movement, physiology, and environment.

02 — LEARN

Represent

Find structure in long time-series and multimodal data at biological scale.

03 — TEST

Interpret

Link patterns to context and candidate meaning with uncertainty visible.

04 — RETURN

Validate

Design safe, measurable feedback to test whether an interpretation holds.

03 / Research frontiers

Begin with what
we can measure.

Each frontier starts with a tractable signal and a real-world reason to understand it better.

01 / Social animalsAcoustic + behavioral

Hear the social world.

Study how calls, movement, and group behavior carry information—from companion animals to whales, dolphins, and other social species.

VOICE · MOVEMENT · SOCIAL CONTEXT
02 / Living systemsElectrical + environmental

Read biological state.

Interpret plant and organism signals where better understanding can improve care, resilience, and resource use.

ELECTRICAL · CHEMICAL · ECOLOGICAL
03 / EcosystemsDistributed intelligence

Map intelligence in nature.

Connect local signals to wider patterns across populations, habitats, and the living networks around them.

POPULATION · HABITAT · NETWORK
04 / Inside the instrument

An observatory
for living signals.

The first interface is not a chatbot. It is a scientific instrument: a way to see patterns, compare signals, and ask better questions.

  • INPUT Continuous multimodal observation 01
  • MODEL Pattern discovery with uncertainty 02
  • OUTPUT Testable hypotheses for researchers 03
05 / The long horizon

One day, the signal may come from somewhere else.

The same discipline that helps us listen to life on Earth may help us recognize intelligence beyond it. We begin with measurable biology. We build toward a wider frontier.

NEAR TERM Living systems, real data, useful science.
LONG TERM A new interface with intelligence in the universe.

The frontier is shared.

Let’s build the instruments
to listen beyond ourselves.

For researchers, builders, and investors drawn to the next great expansion of intelligence.