In reaching a scientific understanding of consciousness and its function in both intelligent creatures and machines, MMCR aims to develop technologies that contribute to industry opportunities
By better understanding how human consciousness interacts with decision-making, attention, memory, perception and social interaction, we offer the opportunity to incorporate these ideas into intelligent systems that can improve business opportunities and quality of life.
By deepening our knowledge on the interplay between human systems and artificial systems, we can forge new technological solutions. We can build better AI systems and better robots with improved accountability and explainability, enabling us to maintain better control of autonomous systems.
Such a perspective would also contribute to improving ethical outcomes as everyday life increasingly crosses over with intelligent, and potentially conscious, machines.
- Artificial general intelligence through artificial consciousness: As artificial intelligence and robotics increasingly surround us with new, autonomous technology, questions come to the fore – is this technology conscious in any way? What kind of intelligence or insight does it possess? And what might it autonomously decide and why? While some of the questions have to do with the technology itself, others are concerned with what consciousness actually is. What would constitute a successful demonstration of artificial consciousness and would we recognise it?
- Ethics of artificial intelligence: The ethical challenges of new AI are vast and largely unexplored. As we develop machine capacities such as AI and robots for human-robot interaction, do we modify the social representation of human behaviours? Will we need to change the way we use machines if they start to exhibit some of the hallmarks of consciousness?
- Neuromorphic computing: The emerging capabilities in artificial intelligence will have more in common with human cognition than with conventional computer logic – emulating the neural structure and operation of the human brain. A neuromorphic model would process and store information in a similar way to the brain and would make computing tasks like image processing faster and less energy-intensive. Only by understanding more about how the brain computes things, will it be possible to realise the full potential of neuromorphic models.
- Neuro-engineering and bionics: Consciousness research will have significant implications on the medical tech industry as we stimulate or modulate conscious experiences through implants such as the bionic eye, as well as the use of neuro-inspired machine learning approaches to study and treat the absences in awareness associated with epilepsy.
- Day-dreaming robot: How does dreaming affect learning and how can we draw that into the world of AI? Will the ability to dream lead to improved learning? Should we be building artificial systems that daydream and simulate hypothetical futures so that they can learn from them?
- Efficient human in the loop ? At present, one of the main requirements for AI systems is to have a ‘human in the loop’, meaning that the artificial systems are not able to function fully autonomously in ethically sensitive contexts. However, it is unclear exactly what that human is providing. Is it a specific skill or simply the assumption that the human will act with a conscience and have the capacity to be punished if things go badly?