Google DeepMind, a United Kingdom-based artificial intelligence company, is widely regarded as one of the pioneers of modern AI.
Led by Demis Hassabis, the company gained global attention for developing computer programs that taught themselves to play chess and later Go, which is often described as the most complex board game in the world. These breakthroughs helped set new standards in machine learning and reinforced the potential of self-learning systems.
In its early years, DeepMind attracted some of the leading minds in artificial intelligence, many of whom were motivated to explore problems that were widely considered unsolvable at the time. Among them were two South Africans, Shakir Mohamed and Ulrich Paquet.
Mohamed studied electrical and information engineering at the University of the Witwatersrand, where he earned several academic distinctions, including the Chancellor’s Medal and the Bernard Price Prize. He completed his degree with distinction in 2005 and briefly worked at Nedbank in credit and risk analysis for retail home loans.
He later received a Commonwealth Scholarship to pursue a doctorate in statistical machine learning at the University of Cambridge. Mohamed has said that during this period, few people understood or recognized the field of machine learning. His doctoral research took four years to complete, after which he moved to Canada to work as a research assistant at the University of British Columbia in a program focused on neural computation and adaptive perception.
Mohamed returned to the UK in 2013 to join DeepMind Technologies. Google acquired the startup for $600 million the following year. After the acquisition, Mohamed was appointed research director, a role in which he continues to work on general-purpose learning algorithms. His research interests include machine learning, Bayesian statistics, deep learning and sociotechnical AI. His work has been cited more than 37,000 times, according to Google Scholar, with the majority of citations recorded since 2020.
Paquet also began his academic journey in South Africa, studying at the University of Pretoria, where he developed an early interest in artificial intelligence. After completing a master’s degree in computer science, he received a Commonwealth Scholarship to study at Cambridge, where he pursued doctoral research in the same field as Mohamed. His thesis focused on probabilistic approaches to supervised and unsupervised machine learning.
After completing his studies, Paquet worked at a company specializing in facial recognition and image similarity technologies before joining Microsoft as a senior researcher. He contributed to the Xbox One launch team, developing backend recommender systems that tailored product suggestions based on user behavior and preferences.
Paquet later joined Vocal IQ, a UK-based startup acquired by Apple in 2015, which led to his appointment as a research manager at Apple. After less than a year, he returned to the UK to join Google DeepMind.
At DeepMind, Paquet worked on AlphaZero, an AI system designed to train itself to master chess, shogi and Go. The system drew praise for its unconventional and creative style of play, including its willingness to sacrifice pieces in ways that challenged traditional strategies.
After becoming close friends during their studies and later collaborating at DeepMind, Mohamed and Paquet co-founded the Deep Learning Indaba to strengthen artificial intelligence research across Africa. The initiative was inspired by their experience attending international AI conferences where they often found themselves among the only African participants.
The Deep Learning Indaba is held annually. The first event was hosted by the University of the Witwatersrand, and the 2025 edition took place in Rwanda.
