AI Report on Southern Africa http://misa.org The focus group discussions took place in Zimbabwe and South Africa with extensive input from experts representing local universities and those from Namibia and Botswana. AI: Definitions (according to Stanford University) 1 Intelligence might be defined as the ability to learn and perform suitable techniques to solve problems and achieve goals appropriate to the context in an uncertain, ever-varying world. A fully pre-programmed factory robot is flexible, accurate, and consistent, but not intelligent. Artificial Intelligence (AI), a term coined by emeritus Stanford Professor John McCarthy in 1955, was defined by him as “the science and engineering of making intelligent machines”. Humans have done much research towards programming machines to behave in a clever way, like playing chess, but today, we emphasise machines that can learn, at least somewhat imitate human intelligence. Machine Learning (ML) is the part of AI that studies how computer agents can improve their perception, knowledge, thinking, or actions based on experience or data. For this, ML draws from computer science, statistics, psychology, neuroscience, economics, and control theory. Narrow AI: Narrow AI is an intelligent system for one thing, e.g., speech or facial recognition. Human-level AI, or Artificial General Intelligence (AGI), seeks broadly intelligent, contextaware machines. It is needed for effective social chatbots or human-robot interaction, which is currently the only type of AI that exists in the real world. It can complete only discrete tasks that computer scientists have programmed it to do, such as playing a specified song on Spotify after “hearing” a voice command (requiring the AI-driven device to decode a series of syllables) through a virtual assistant such as Alexa or Siri. General AI is what many people envision when they think of AI—machines that exhibit real human intelligence and can feel, innovate, or emote, like HAL 9000 from the Space Odyssey series or Samantha from the film Her. Outside such fictional realms, scientists have yet to code human emotions, awareness, and consciousness into machines. Super AI is a step above general AI and is defined as AI that surpasses human capacity. Again, super AI is a theoretical concept, yet to be achieved. Deep Learning uses large multi-layer (artificial) neural networks that compute with continuous (real number) representations, a little like the hierarchically organised neurons in human brains. It is currently the most successful ML approach, usable for all types of ML, with better generalisation from small data and better scaling to big data and compute budgets. An algorithm lists the precise steps to take, such as a person writes in a computer program. AI systems contain algorithms, but often just for a few parts, like a learning or reward calculation method. Much of their behavior emerges via learning from data or experience, a sea change in system design that Stanford alumnus Andrej Karpathy dubbed Software 2.0. 1 AI definitions https://hai.stanford.edu/sites/default/files/2020-09/AI-Definitions-HAI.pdf 6