Our Work

Fundamental Science

• Fundamental and foundational AI research  
• Capability development; an interdependent, productive network 
• Advancing cutting-edge research and technology innovation in AI   
• Defining strategic areas of research, mobilizing resources, and cultivating a collaborative environment that drives scientific discovery and innovation   

Uniting interdisciplinary teams; facilitating productive collaboration This area of MIND’s work seeks to facilitate tangible, outcomes-driven collaboration, through multi-disciplinary engagement model. Programme activities provide platforms for identification, problem-solving and collaboration, looking for new intersections of research? 

This work relates specifically to the enhancement and development of new models and algorithms in domains such as autonomous decision making and planning, computer vision, language and other foundation models, and probabilistic models. The core focus of this work is on improved performance, scalability, data and compute efficiency, and safety and explainability? 

?Biologically-inspired information processing:? 

MIND’s unique composition will allow for a strong focus on developing biologically-inspired learning mechanisms. This can be achieved by leveraging insights from neuroscience and anatomy to developing machine learning approaches that mimic biological processes? 

Work in this area will also focus on using insights from psychology and cognitive science to shape improved representation learning techniques that are informed by human and animal perception and knowledge acquisition strategies. Such representations would result in AI models better capable of generalising and reasoning in complex environments? 

As a major AI institute in Africa, MIND is positioned to lead development of energy- and data-efficient AI models, addressing both local needs and challenges in developing countries. Research will draw on Africa's experience with resource-constrained environments, with a focus on models that perform effectively with limited computational power and sparse data. Inspired by evolutionary efficiency in biological systems, we aim to develop algorithms optimised for low energy use, making AI more accessible in remote and underserved areas? 

Through its multidisciplinary approach, MIND is ideally positioned to use AI to probe and validate hypotheses about human and animal cognition. By building appropriate analogues, we can test theories of intelligence and cognition. This approach not only fosters AI advancements inspired by biological principles but also drives new discoveries in biology and medicine, potentially leading to breakthroughs in understanding cognitive mechanisms? 

Developing novel architectures for low-resource, highly diverse settings? 

Africa is an ideal setting for exploring multilingual opportunities – its rich linguistic diversity provides a unique foundation for developing advanced multilingual models. By creating models informed by the continent’s various languages and dialects, we can capture the cognitive and cultural nuances that shape communication. This research goes beyond merely developing multilingual systems; it also allows us to investigate how multilinguality influences human cognition, opening avenues for understanding cognitive processes.

Application

• Collaborating on applications to transform advances in AI into responsible, market-ready innovations for sustainable impact   
• Fostering the incubation and commercialization of innovative technologies in collaboration with industry, including large-scale AI models applicable to various industries  

MIND aims to collaborate effectively with industry and social development actors towards effecting sustained and systemic impact across economic as well as other development challenges. 

AI Governance

Defining AI policy and legislative contributions and innovations locally and globally considering the African socio-economic imperatives, sustainable development needs and goals, culture, and diversity  

Enabling assessment, development and deployment of human-centered AI, designed to benefit humanity (and our biosphere), and enhance human capabilities  

Strengthen context-relevant governance insight regarding ethical AI development and use, security, safeguards and consideration of existential risks  

Advocacy around key African perspectives on AI digital diplomacy and geopolitics; digital rights and inclusion; trust, accountability and transparency.