Gird your loins - the robots are coming to Africa

Leading neuroscientists, robotics and data scientists share their thoughts on what AI and robotics mean for the continent.

By Iga Motylska

Pictured: Ashley Anthony, CEO of Isazi Consulting

One of the foremost questions about artificial intelligence (AI) and robotics is the likelihood of humanlike and self-aware robots with thoughts and feelings of their own, similar to the ones we see on television. “There is no such thing right now. We have ideas of what to do, but deep neural networks – which are the in-vogue technology right now and which is incredibly transformative – will not lead to artificial general intelligence,” said Vivienne Ming, from the Singularity University Faculty of Cognitive Neuroscience during a panel discussion on robotics and AI. “It’s not impossible, but we need a new paradigm.”

Sarah Bergbreiter, from the Singularity University Faculty of Robotics, echoed this sentiment, saying: “The challenges are also on the hardware side of robotics. Right now, robotics hardware is designed for a specific task. We design it to do one thing incredibly well, but it’s hard to design something in hardware that can do a number of things very well.”

Ming says we are too caught up in metaphors of intelligence that are similar to us. Robots and AI are largely created for our self-interest and benefit. She said:

“Genuinely intelligent systems will be a big distributed system, that because of the way it is embodied, because of the time scale it operates on, has nothing to say to us and we have nothing to say to it. These are not going to be robots that we can have conversations with.”

However, she believes there’s great potential in the sphere of effective computing, which can tap into our emotional states – not mimic our feelings – in ways that we aren’t even aware of. “There’s the possibility to make systems that will emulate emotional systems richly enough, whereby we won’t be able to distinguish [whether the robot is conscious or not because the feedback and responsiveness will trigger an intuition as if it is conscious], that should be quite scary,” said Ming.

Bergbreiter added that there’s the possibility of instrumenting factory workers with EMG (electromyography is an electrodiagnostic technique to record electrical activity produced by skeletal muscles) to know that they are about to move; however, with exoskeleton robots it’s still easier to train the human to the robot than to train the robot to the human.

“I am very uncertain of the future of artificial general intelligence, however, the future augmented intelligence in terms of neural prosthetics using micro robots is certain,” said Ming. She believes that in the next 30 years, these technologies will fundamentally change the definition of what it means to be human.

On the topic of automation and mechanisation, Ming noted that trying to regulate the technology side of things is a case of considering the wrong aspect of the problem because it’s a human problem that requires social institutions to keep pace with this change, instead of being entrenched with the way things have been run in the past.

On the continent, the practical everyday applications of such technologies can be seen through work done by Ashley Anthony, CEO of Isazi Consulting. He said:

“Our goal is to define businesses in Africa mathematically and to look at technologies in AI with a focus on machine learning and optimisation. We apply these various technologies to solve every day problems. From a science perspective, we can think of it agnostically and pull down the walls between applied statistics, applied and pure mathematics, physics, electrical engineering and computer science, and really have the ability to jump into one, pull it and apply it to any problem.”  

Isazi constantly asks how AI and optimisation can assist companies to become more data driven so as not to become redundant. “Solving an optimisation problem is important across Africa and also speaks to the pure fundamentals of any business,” Anthony said.

He believes AI will help even out the playing field for developing countries, such as South Africa. This approach can help societies be more practical and answer questions such as how to optimally distribute fibre to help create a more connected continent, as well as how to optimally distribute a business’ capital based on analysing its data trends. “Without serious investment into AI from both private sector and government, we will be left behind. South Africa has a habit of exporting and buying back our solutions for our problems,” he said.