Adaptive Intelligence With Digital Brains

AI that learns, understands & interacts with the world like the brain

Our mission

We build AI with Adaptive Intelligence that can learn, understand, and interact with the world like the brain to make intelligent interactive real time assistance accessible to everybody.

The fastest learning AI.
Masters Pong in under 5 minutes on a laptop CPU.

Wins after a few games. Runs on commodity hardware. Over 100x more efficient than leading reinforcement learning models.

See how this changes robotics and real-time AI → Technology

The scaling problem

For brains, prior learning makes new learning easier. For today's AI, it's the opposite—each new capability costs more data, compute, and time. It can't accumulate enough skills for general intelligence. That's why it won't work for robotics.

Solution: Causal AI understanding cause and effect like us

Causal AI is a powerful paradigm for physical AI, because it learns the causal rules that generate the environment and can therefore, reproduce it, extend it into the future, reason about it, and act on it – causally and in real-time

Our uniqueness

With unmatched expertise and experience in developing one-of-a-kind Digital Brain technology, we are building AI for life-like interactions in complex, dynamic environments. Our revolutionary approach involves teaching digital brains, the closest proxy to real brains, to naturally evolve cognitive skills. By leveraging modern AI for sensory pre-processing and for language, we are building inait Adaptive Machines (iAMs) to provide holistic, generalizable AI for all industry verticals, starting with finance and robotics. With key strategic partners in technology & go-to-market and with robust financial backing, Inait represents a winner-take-all opportunity in the post-genAI era.

Modern AI cannot

achieve life-like interactions in dynamic physical, virtual and data-rich environments


accumulate new skills without an exponential rise in data and energy needs


develop exponential cognition that is needed to reach generalized forms of artificial intelligence


keep up in dynamic environments because correlations are insufficient to learn the rules of engagement


grow in intelligence remotely, relying on expensive centralized version updates

AI with Digital Brains learns

cause and effect from actions and consequences, allowing learning of the rules of engagement and the natural emergence of cognitive skills


new skills easier and easier as seen in all biological organisms, not harder and harder as seen in all modern AI


to accumulate the full cognitive skill sets needed to adapt to interact intelligently in any given dynamic environment


and develops agency through embodiment to achieve life-like interactions in complex dynamic environments


continuously and remotely from new experiences without needing retraining with more data and energy

The Founder

Henry Markram
Professor of Neuroscience and serial entrepreneur

Henry has pioneered groundbreaking initiatives that have redefined brain research and computational neuroscience, including founding the Brain Mind Institute, the Blue Brain Project, and the Human Brain Project, and serving as co-founder and chairman of Frontiers, the Frontiers Research Foundation, the Frontiers Planet Prize, and Inait. He is also the founder and president of the Board of the Open Brain Institute and Inait. A visionary leader and pioneer in simulation neuroscience, Henry has led revolutionary efforts to work out how to build digital brains, how they work, and how they learn.

The Launch

The Dawn of Digital Brains

At the recent World Economic Forum, Markram presented a five-step recipe for building digital brains: populate them with neurons, grow dendrites, grow axons, form synapses, and model the electrochemical behavior of each neuronal and synaptic type—then “switch on” these networks to iteratively refine their biological fidelity. He further explained how teaching these digital brains can revolutionize AI by mirroring the brain’s own adaptive capabilities, opening the door to a future where generalizable AI arises from genuine causal understanding rather than mere correlation-based methods.