A research lab · Europe
Intelligence is not enough. AI needs a way of being.
Humaneness Labs is building the behavior layer for AI — models of attention, emotion, perception, and social context that make interaction feel present.
it breathes. move your cursor closer.
01 · The missing layer
Capable, expressive —
and still not quite there.
AI can speak fluently, read a room, generate a face, and move through the physical world. Yet even the strongest systems often feel subtly performed: the voice that is too polished, the pause that arrives on cue, the expression without a thought behind it.
The gap is not more expression. It is behavior with something beneath it.
02 · Beneath behavior
We are never
only one thing.
Human behavior does not come from a single emotional label. It emerges from many streams moving at once: attention, emotion, memory, anticipation, perception, physical sensation, social context, and thought.
We can be warm while tired, amused while distracted, curious while uncertain. We read a voice, a face, a room, and our own body at the same time. Thoughts interfere. Attention wanders. We hesitate, recover, and change our minds.
These imperfections are not noise. They are part of what makes behavior feel alive.
03 · The first wedge
Start where the difference
can be heard.
We are beginning with voice because it reveals far more than words. Timing, breath, rhythm, effort, hesitation, interruption, and tiny shifts in articulation all expose what is happening beneath the surface.
Rather than adding emotion after the fact, we are researching systems in which expressive behavior emerges from context and changing internal state.
A voice with an inner life
04 · One behavior layer
Voice is the beginning,
not the boundary.
Speech shaped by context, attention, emotion, and interruption.
Perception of gaze, posture, expression, gesture, and social signals.
Reactions that unfold from state instead of playing as animation.
Digital beings with coherent attention, memory, and presence.
Movement and interaction that respond fluidly to people and place.
The long-term system connects perception, internal state, and action in one continuous loop.
05 · The direction
From simulation
to presence.
Our goal is not to make machines better at pretending to be human. It is to understand the computational principles that make interaction feel authentic — and build systems that can use them in context.
From imitation toward understanding.
From performance toward presence.
From realism toward genuine interaction.
We are building the soul
of the machines of the future.
Research partnerships