Nirvanic’s ‘Spark of Life’ Demo Puts Quantum Chips in Control of an AI Robot at MARS 2025

Nirvanic Consciousness Technologies delivered a deceptively modest robotics dEMOnstration at Jeff Bezos’ private MARS 2025 conference, yet the event may be remembered as a landmark moment for both AI robotics and the study of consciousness. The company stresses that its participation does not imply any endorsement from Bezos or Amazon.

Nirvanic is a legitimate deep-tech startup founded by Suzanne Gildert, a British-born Canadian regarded as one of the most extraordinary innovators in recent Technology history. Gildert holds a Ph.D. in Experimental quantum computing, spent years developing quantum AI algorithms at D-Wave, co-founded Kindred AI (which achieved the third-largest exit in Canadian robotics startup history, at CAD 330 million), and later co-founded Sanctuary AI, where she served as CTO during the development of the advanced humanoid robot Phoenix. In 2024, she left Sanctuary to found Nirvanic with an audacious mission: “to engineer consciousness technologies.”

Gildert believes the current technological landscape offers an unprecedented opportunity to test the theory of quantum consciousness—the hypothesis that conscious experience arises from quantum effects such as entanglement and superposition in the brain. While this decades-old theory has long been dismissed as pseudoscience, recent experiments have provided evidence that such quantum processes may indeed occur in the wArm, wet environment of the human brain, lending the idea new credibility. Gildert argues that sentience is the missing ingredient for the next major leap in AI robotics: the ability to shift from unconscious, pattern-matching behavior into a creative, aware problem-solving mode when encountering situations far outside training data.

At Mars 2025, Nirvanic conducted its foundational “Spark of Life” experiment. A small quadruped robot equipped with a webcam sent perception data to a distant D-Wave quantum computer on Canada’s west coast. Quantum code processed the input, collapsed the wave function, and returned a chosen ACTion from 32 possible options—encoded by 5 qubits. The robot then moved, completing this loop roughly twice per second. Gildert emphasized that the robot is not claimed to be conscious; rather, the platform is designed to search for behavioral signatures that distinguish quantum-controlled decisions from classical ones.

The research program now underway involves running identical setups in Classical simulation and genuine quantum modes, collecting massive datasets of action choices, and comparing statistical distributions. If the quantum system consistently exhibits distinct behavioral patterns, the next step will be to amplify those quirks through reinforcement learning. Should learning accelerate with a quantum component in the loop, it could evolve into a transformative technology. Gildert also suggests that if consciousness proves to be a quantum phenomenon, it would provide crucial ethical clarity, creating a recognizable boundary between conscious and unconscious states in future sentient machines.

Nirvanic’s long-term ambition is twofold: to validate or falsify the theory of quantum consciousness through rigorous experimentation, and to develop commercial quantum computing tools that significantly advance AI and robotics. Gildert likens the current state of quantum computing to the early days of GPUs—powerful hardware still waiting for its killer APP. Consciousness research, she proposes, just might be that application.