New Paper Maps Route for the Future of AI

VERSES Publishes New Research Proposing Future Path for AI

White paper describes novel approach to machine intelligence that underpins VERSES platform

VERSES announced that a team of fifteen of the world’s leading Computational Neuroscientists and AI Ph.D.’s, in association with Universities across Canada, UK, Germany, Australia, The Netherlands, and the USA, published a landmark research “white” paper describing their novel approach to the design and development of the next generation of AI known as Artificial General Intelligence (AGI) intelligent agents with human-level intelligence that can learn, adapt and act in the world. The paper, entitled “Designing Ecosystems of Intelligence from First Principles,” presents a vision and roadmap for AI based on a field of research called Active Inference. KOSM, the network operating system from VERSES, uses the Active Inference framework as the foundation for its AI decision-making and planning capabilities.

Artificial Intelligence has broad applications spanning industrial automation, shopping recommendations, computer vision, voice assistants, self-driving vehicles, generative art, customer service, and more. The global AI market is projected to grow to USD $1.394 trillion by the end of the decade. Despite its successful applications, AI has fundamental challenges impeding the ultimate aspiration of human (or super-human) level generalized machine intelligence. Conventional AI works by recognizing patterns within large data sets that can cost millions of dollars to train and deploy. Moreover, today’s AI models are custom-built for narrow problem sets, for example, a model designed to alleviate traffic congestion can’t be applied to help identify medical issues in x-rays.

VERSES Chief Scientist, Dr. Karl Friston, pioneered a different approach modeled on the way that humans – and things in nature more generally – understand the world and learn patterns. This approach, called Active Inference, describes how biological intelligence observes the world, orients, and makes decisions about how to act based on sensory observations or data, all in real time in contrast to today’s AI known as Deep Learning which heavily relies upon massive historical data sets. Active Inference acknowledges that the world is a messy place and that fundamental to great decision-making is reducing uncertainty in order to achieve expected outcomes. “Active Inference is a universal scheme for intelligent behavior. Operationally, Active Inference is a process of belief updating and propagation that has two aspects: inferring states of affairs in the world that generate data and inferring the best plans for actively acquiring more data,” said Friston.

VERSES CEO, Gabriel René, comments, “We are moving from the Information Age to the Intelligence Age, and this groundbreaking paper – the culmination of decades of advanced scientific research and development – describes, for the first time, a genuine path to achieving the ‘holy grail’ AGI not as a single algorithm but as an ecosystem of collective intelligence. This powerful science is the foundation of our KOSM platform.” Using Active Inference, KOSM enables the design and deployment of a new generation of intelligent agents capable of working together to provide unprecedented insights, recommendations, and autonomous capabilities to enhance any application across any industry. René continues, “VERSES mission is to enable a smarter world where the power to design and deploy AI is extended beyond those with big data to anyone with big ideas.”

VERSES Research Lab collaborated with the following universities on the white paper:

Wellcome Centre for Human Neuroimaging, University College London, London, UK,
Department of Philosophy, Monash University, Melbourne, Australia Sussex AI Group
Department of Informatics, University of Sussex, Brighton, UK
Department of Computer Science, Université du Québec à Montréal, Montréal, Québec, Canada Department of Experimental Psychology, University of Oxford, Oxford, UK
Department of Collective Behaviour, Max Planck Institute of Animal Behavior, Konstanz, Germany Department of Biology, University of Konstanz, Konstanz, Germany
Centre for the Advanced Study of Collective Behaviour, University of Konstanz, Konstanz, Germany
Network Science Institute, Northeastern University, Boston, Massachusetts, USA
Brain Network Dynamics Unit, University of Oxford, Oxford, UK
Department of Mathematics, Stony Brook University, Stony Brook, New York, USA
Department of Physics and Astronomy, Stony Brook University, Stony Brook, New York, USA Department of Biomedical Engineering, Stony Brook University, Stony Brook, New York, USA Topos Institute, Berkeley, California, USA
Department of Electrical Engineering, Eindhoven University of Technology, Eindhoven, The Netherlands
Department of Philosophy, Université du Québec à Montréal, Montréal, Québec, Canada

The contextualized data structure and messaging that KOSM employs are based on open specifications developed by the Spatial Web Foundation and being standardized within the IEEE Standards Association. The Designing Ecosystems of Intelligence from First Principles white paper is publicly available at arxiv.org.

Additionally, VERSES announces that it has granted an aggregate of 400,000 stock options to certain consultants pursuant to the Company's stock option plan. All of the stock options will be exercisable to purchase one Class A Share at a price of $0.75 for a period of 1 year from the date of grant. The options will vest immediately upon the date of grant.

Source: VERSES media announcement

Explore Cybersecurity Solutions for ISPs

Follow @PipelineWire