The IndabaX machine learning and artificial intelligence conference was hosted at UKZN. The opening discussion on AI from an African perspective contextualized the gathering and adjusted my mindset to unique problems data scientists face on the continent. The opening keynote presentation by Prof Tshilidzi Marwala presented the history of industrial revolutions and how the rationality of artificial machines will contribute to the evolution of the AI field. Prof Marwala is currently the Vice-Chancellor and Principal of the University of Johannesburg and has authored over 10 books on artificial intelligence..
IndabaX
The second keynote by Dr Andrew Saxe on high-dimensional dynamics of generalization error in neural networks admittedly went over my head, but this only motivated me to expand my knowledge on neural networks. The conference in general showed me not only what is possible but also what is left for me to learn. The final keynote by Prof Bruce Bassett was, in my opinion, the best of the conference. His talk genetic algorithms and how they can be used to generate the best possible learning infrastructure was absolutely mind blowing. Although some of the finer details were too complex for me at the time, the notion that a machine could not only restructure it’s decision making process but also optimise it’s own system for a given task lead me to believe that artificial general intelligence (AGI) is possible within our lifetime. This idea is both incredible and terrifying at the same time.
One of the research presentations which caught my attention was by Dr Emli-Mari Nel on Ticker: An Adaptive Single-Switch Text Entry Method for Visually Impaired User. Ticker allows patients with locked-in syndrome to communicate by listening the the alphabet being spoken by 5 different voices simultaneously and selecting a desired letter by making a predefined eye or facial muscle movement if possible. Another interesting research presentation was that by Dr Maria Schuld on doing machine learning on quantum computers.
