Page 1


4th QUARTER 2018 ZAR25 | US$2.50 | Euro1.60

The Voice of African AI & Data Science











AI will Free your Employees from the Doldrums

6 A new Paradigm for Building AI for Good, for Africa 8

Building African Youth Today for the AI Future

10 AI Expo 2018 in Pictures 16 Personal AI 18 Mark the Date - AI Expo 2019 Prospectus 20 What Delegates Had to Say about Africa’s Inaugural AI Expo 21 Survey Results - AI Expo 2018 24 Award Winners 2018 Expo 26 Zindi & MIIA Partner to Transform Africa through Data Science 26 How to Address the Massive Shortage of Data Science Skills in Africa and the World 28 The Good, the Bad and the Ugly of Applying Ai in Customer Services 29 How AI is Transforming Health Care in Africa 31 Pathways for the Adoption of AI by Managers 33 The Potential of AI in Africa 35 The Real Impact of AI on Jobs 39 Ethical Considerations for Artificial Intelligence 40 Intel Movidius Neural Compute Stick Winner 41 DJI Mavic Air Drone Winner

WHEN NICK and I set out upon this journey of creating a platform for Africa’s fledgling AI & Data Science community to showcase its talents and potential, and unify the community under a single umbrella with the shared vision of harnessing AI for the greater economic and social good, we were delighted to find a continent rich in innovation, AI visionaries and pioneers. We were even more delighted to see the community rally around our event - AI Expo Africa - especially since no-one had ever tried anything similar, and in many instances, more than a few failed to see our vision. The support we received for our first event was overwhelming, and for those visionaries that joined us - we say a heartfelt thank you, and we hope to see you all at next year’s AI Expo 2019, as we bring an expanded event. So mark the date: 4 & 5 September 2019 - same place (Century City Conference Centre), same great food and atmosphere and an expanded expo floor, with expanded conference and training offerings. Click on the showreel below to get a recap of AI Expo 2018, and watch our various social media channels for early bird ticket ticket offerings. Sponsorhips are now open too - please e-mail: Roy Bannister - AI Media Co-Founder & Editor-in-Chief

THEY SAY Rome wasn’t built in a day, and that’s exactly how we felt about the lead up to AI Expo Africa and the launch of Synapse magazine. We worked on the vision, concept and execution of Africa’s largest business focused AI community event for over a year and our September 2018 launch turned out to be an outstanding success. This edition of Synapse hopes to capture the spirit of the event and feedback for those that missed it as well as share new content from contributors across the region. Thanks again to all those that supported us and took a leap of faith - we truly hope to build on this momentum and grow the African AI & Data Science community across the continent in the months and years to come. Nick Bradshaw - AI Media Co-Founder & Community Director

AI Is Going To Free Your Employees From The


Knowledge workers are plagued by a good deal of mind-numbing drudgery. And the effects are harmful. Busy Work Is Causing Widespread Professional Dissatisfaction


By Alex Terry - CEO, Conversica



CONVERSICA RECENTLY polled more than 1,000 knowledge workers to find out how much of their work day is spent on routine, unimaginative and repetitive tasks, as well as to better understand the toll this drudgery is taking on their emotional and mental health. The results were illuminating. More than half say their day involves a significant amount of “busy work” — 42 percent spend 30 minutes of each hour on it — and 80 percent say aspects of their job are beneath their skill level. When asked to assign a letter grade to their workplace – from “A” for highly engaging to “F” for soulcrushing – 36 percent gave their workplace a grade of “C” or lower. Drudgery is simply terrible for morale. Survey respondents said this “busy work” makes them feel dissatisfied, bored, discouraged, stressed, and undervalued. Almost two-thirds said a job heavy on drudgery makes them feel like they are wasting their lives. This is also terrible for employers: in economies where there are more job openings than job-seekers, more than two-thirds of workers cite drudgery as a reason for leaving their job. Most Believe AI Will Be Able to Help Workers are ready for more technological innovation and intervention. According to the study, about half believe that some or all of their busy work could and should be automated, and more than half believe that some of their routine work conversations could be automated. Two-thirds expect to be using AI for this purpose within the year, and three-quarters expect to be using it within five years. This is a logical assumption, considering how much AI

and automation have already reshaped the workplace. Conversational AI advances, using natural language processing and machine learning, allow human-human interactions to be automated. Like your best employees, conversational AI assisants write, read and understand email and text messages and take the right actions, freeing up employees to work on the cognitive and creative tasks that make them happier. Even better, these AI assistants are infinitely scalable, work 24x7, and never tire of the routine. Indeed, as a recent article by Deloitte in the Wall Street Journal titled “The Evolution of Work” points out: “the falling cost of automation, an increase in the use of AI, and the rise of human-machine collaboration have created a new reality of talent trade-offs and transitions. This presents the opportunity to reimagine the economic value of work through the increased productivity that human-machine collaboration can bring to the workplace.” This is great news for workers — and employers. Freed from busy work, workers report that they would focus on more rewarding work, be motivated to get more accomplished each day, showcase their true talents and abilities, and even have better personal relationships. In other words, they’d be happier. Goodbye Drudgery, Hello Delight Today we possess the technology to eliminate more and more rote tasks from our employees’ lives. Conversational AI in particular is helping to eliminate some of that drudgery that keeps our employees from delivering their best. Now that such automation is available, we have the obligation to put it into action, since less drudgery means delighted employees, and delighted employees help us created delighted customers. ai

Alex Terry is the CEO at Conversica, a technology company that provides conversational AI solutions to help businesses automate important yet routine human interactions.

Conversational AI Assistants


AI Assistant

Monday 9:02

Hi David,

I saw you downloaded information from our website. Can I provide you with any additional information to better understand our product? Best regards, Michelle



Tuesday 23:17

Hi Michelle,

Thank you for the quick reply. Yes, I do have some questions about your product. I am free Thursday. Best, David

Hire an AI Assistant to engage current and future customers Conversica’s AI Assistants leverage advanced AI technologies and free your team to focus on what they excel at: relationship building •

Engages in 2-way high-quality personalised conversations

Communicates through email, SMS, and other text-based channels

Wednesday 9:03

No, that's my office number. I am working from home today so please call my mobile 011-625-5555. Thank you for burning the midnight oil on my behalf.

Converses in multiple languages and at consistently high quality

Always learning - Gets smarter with every engagement


AI Assistant

Tuesday 23:26

Hi David,

Our sales manager Michael can help you and would like to give you a call. Is +27 11 978 1234 the best number to reach you on during the day? Many thanks, Michelle

From Hi Michelle,


Best, David


Wednesday 10:12

Hi David,

AI Assistant

Automate and personalise customer conversations at scale

No problem. Michael will call you this coming Thursday.

Go from drudgery to delight

Hire an AI Assistant for your organization today!

All the best, Michelle

Visit to learn more © 2018 Conversica, Inc. All rights reserved.

A new paradigm for building AI for good, for Africa September was a great month for the advancement of artificial intelligence (AI) in Africa for businesses, innovators and students. I facilitated, participated in and listened to three panel discussions on AI on (and for) the African continent at AI Expo Africa, the Innovation Summit and the second Deep Learning Indaba. Given the enthusiasm that currently surrounds AI technologies, these conversations have prompted me to reflect deeply on the implications of AI ethics in the African context.


Data in itself has little value. To unlock the value in data, data mining techniques are required to extract information and gain insights.



The notion that data is the “new gold” dominates business conversations revolving around data science and AI. Reminiscent of the gold rush of the mid 19th century, many aspiring digital companies gather consumer data with insatiable hunger. Data in itself, however, has little value. To unlock the value in data, data mining techniques are applied to extract information and gain insights. Many conversations around AI ethics focus purely on whether insights are used for good or evil - what types of insights are gained, and what contexts they are applied to. Using AI for social good could, for example, imply extracting insights from data to improve health outcomes, increase access to finance or improve the safety of groups of people, particularly in environments where economically impoverished people suffer due to a lack of resources. While I fully support the application of data and AI for good, as well as accelerated global progress towards the sustainable development goals, that alone is not enough. The fundamentals of business lie in solving problems that people care to have solved. Each problem that is solved has economic value and the potential to generate wealth. The tech sector has a history of distributing its wealth to a very small number of individuals, leaving most of the world’s impoverished population with no option but to accept ‘free’ products. While a free solution may provide temporary relief to a beneficiary, it is the process of beneficiation that provides long term gain to the solution provider. Fair only, if all stakeholders that contribute to creating the solution are compensated for their value-add.  Anybody who is familiar with Africa’s history ought to feel a sense of unease at the language used in the data world: gold, mining, extraction. Despite the potentially exponential impacts of AI technology, anybody who is familiar with Africa’s history ought to feel a sense of unease at the language used in the data world: gold, mining, extraction. Africa has a long and complex history of

by Wiebke Toussaint Applied AI Researcher & Entrepreneur

resource extraction and plundering. Over the past centuries very few companies have fairly compensated African people for the mineral and natural wealth they have extracted. Moreover, local communities often carry the environmental burden of waste, resource depletion and pollution, as well as social costs associated with a loss of traditional lifestyles. While the resource extraction sector still has a long road ahead towards sustainability, some mining companies have started recognising the importance of having a social license to operate, with themes such as “mining for shared value” becoming increasingly important. When contemplating data as the fuel for AI technologies and the 4th industrial revolution, it is useful to apply this resource framework to unlock key questions that are missing in the AI ethics narrative. What are the hidden costs in the world of data? Who carries the burden of these costs, and who pockets the gains? Do individuals from vulnerable groups understand the future cost of divulging personal information and data that capture behavioural patterns? Are local communities treated merely as data (read resource) providers, supplying their data (read gold) free of charge to foreign companies that do offshore beneficiation to sell value-added products back to Africa?  We need to move away from seeing data as the new gold, and instead commit to systems that use data for shared value. Yes, we need better health care, energy access, agricultural produce, education, safety, water. Yes, we also need equal opportunity to attain economic freedom. Receiving data products without the proceeds of the beneficiation process being returned to Africa will render the data revolution into an exponentially extractive industry that perpetuates a global history of exploitation of the world’s most vulnerable people. If we want to create a future that is better than our past, individuals and organisations doing “AI for Africa” need to be mindful to leverage the data revolution for the gain of all people on the continent. We need to move away from seeing data as the new gold, and instead commit to systems that use data for shared value. ai

Building African youth today for the AI future Speaking in an address to commemorate the 200th anniversary of the foundation of the University of Pennsylvania in September 1940, the 32nd president of the USA, Franklin D. Roosevelt, commented: “We cannot always build the future for our youth, but we can build our youth for the future.”


By Jake Mwangi-Powell, International Baccalaureate student at Braeburn Garden Estate School, Nairobi, Kenya THIS SENTIMENT certainly resonates today in discussions concerning Artificial Intelligence (AI) and youth globally, but is particularly prescient for African youth. Constituting under a fifth (1.28 billion) of the world populace in 2018, 41% of Africa’s population is aged under 15 years; while this figure will decline to 32% by 2050, there will be absolute increase in demographic numbers.1 Moreover, with 200 million people aged between 15-24, Africa has the largest population of young people in the world. Neglecting the involvement of this potentially talented and creative age group in meaningful employment risks social unrest, what Zambia’s former finance minister, Alexander Chikwanda, described as “a ticking time bomb.”2 But omitting them from new, future technologically innovative landscapes created by AI could also seal Africa’s fate as a passive consumer rather than active and essential producer of such innovations. So, how do we ‘build youth’ for this new future? Clearly, as I outline below, education is central to this agenda. But there are other factors, too, including general awareness-raising to make the reality of AI concrete, understandable and approachable, rather than being what for many (adults and youth included) is abstract and unknowable or a dystopian vision of the point of Singularity produced by Hollywood studios. Education is, however, critical, but it is not just educational opportunities at the university level. There is a need for quality education,

inspiration for what is possible and a celebration of enquiry and academic achievement across all educational stages. For example, these could include: Primary and secondary level: Focusing on STEM (science, technology, engineering and mathematics) subjects, the formation of STEM and AI clubs and inter-school competitions to celebrate success, and the promotion and use of youth champions for AI to educate and energise pupils within schools on STEM, robotics and their possibilities. University – undergraduate: Emphasizing and investing – via public-private partnership scholarships – in foundational mathematics, engineering, computer science, economics and medicine that have AI embedded within them on a modular basis. University – postgraduate: Building research capacity with scholarships that favour AI exploration; creating hands-on work experience opportunities with African start-up organisations and international organisations where they do not yet exist. Post-university: Investing in commercially supported startup opportunities for graduates. It is anticipated that by such strategic and coordinated interventions, African nations will be able to fulfil President Roosevelt’s recommendation and build its youth for the AI future. ai


8 REFERENCES 1. Population Reference Bureau. World Population Data, 2018. Accessed: 19 October, 2018. 2. Anon. How 5 of the most creative young social entrepreneurs are addressing youth employment in Africa. All Africa, 13 January, 2015. https://allafrica. com/stories/201501271701.html. Accessed: 19 October, 2018.

Accelerating Digital Transformation

Impactful Decision Making

Reinvent Problem Solving

Unmatched Customer Experience

Intelligent Devices

Smarter Execution

Total Cybersecurity

To Schedule a Demo on Mosaic, Contact E: | T: 0113179200 | W:


Continuous learning



in Pictures

















Personal A.I. Engaging with customers daily can sometimes be fascinating. With so many new ideas and innovations in the tech world every year, the eagerness to talk to them and share with them is great. It can be humbling then, astonishing even, when you go to a customer to talk about a topic like Machine Learning, only to see that not only have they embraced your technology, but are already pushing the boundaries in ways you haven’t seen previously. And when the person giving you a detailed instruction of what they’re doing is barely out of college, it does fill you with some excitement and optimism. The gist of the above is that terms like Machine Learning and A.I have entered the mainstream lexicon very quickly in recent years, and currently it is fairly normal to talk about these concepts outside the tech sphere.


THE CLOUD has really enabled us to both speed up development in the A.I space (note how this field has progressed post-cloud), and bring A.I technologies to a wider audience with the cloud’s consumptive model. A big advantage of the cloud, specifically for Machine Learning, is gaining access to hardware platforms that you perhaps wouldn’t have invested in previously in the on-premises world. Azure, for example, gives you access to GPUbased VMs, on which you can build HPC and AI solutions, scaling quickly to even 1000s of GPUs. In 2017, Microsoft revealed Project Brainwave, a platform built on FPGAs (Field Programmable Gate Arrays) in partnership with Intel. This is a leap over simply chaining GPUs together, the entire architecture is optimized for Deep Learning. At the unveiling of Brainwave, the Intel Stratix 10 FPGA, built on a 14nm process, demonstrated performance of a sustained 39.5 Teraflops. These hardware innovations continue to be delivered, and will continue to improve society’s Big Data and Machine Learning capability in the cloud. But for the average person, this will be invisible, it will run in the background, and they won’t even be aware that all of this is improving their lives. To complete the picture, we need to move from the cloud to the edge…..



The Intelligent Edge In the world of IOT a new concept is becoming prevalent – the Intelligent Edge. Edge computing refers to data processing power at the edge of a network instead of holding that processing power in a cloud. Now this may seem odd to some of you, however the importance of the cloud does not diminish in this scenario. The intelligent edge recognizes that to deliver what businesses require, data processing and intelligence sometimes needs to be applied at the edge before data is synced into the cloud.

by Thavash Govender Data & AI Solution Specialist at Microsoft

The business world has talked this IOT scenario for many years, and not only have we seen many examples of potential solutions being proposed, we can now say that in certain fields like manufacturing and mining, IOT solutions are a reality and are already delivering business value. But what about the world away from business and production plants ? How will this “Intelligent Edge” technology touch the average human being in his or her daily life in the near future ? This is the concept that has fascinated me for years. I have always seen AI as being personal, something that ultimately will enhance the capabilities of human beings in business, work and life. In the 20th century there was a time when a computer was generally accepted to be a large, room filling device, mysterious to the normal person although they were aware that it existed and could perform amazing tasks. This is exactly what Artificial Intelligence is today. However, in the late 1970s, as technology allowed , a computer could be made that was relatively much smaller , designed for a single user and to enhance the capability of that user. This will also happen with Artificial Intelligence. Many years ago, in my various blogs, I theorised about the concept of having tremendous processing power on you, enough to power AI. In 2010, I spoke of the concept of a “Local AI”, tremendous computing and analytical power in your pocket ( and on your wrist ) , that augmented your capabilities and changed your daily life. In 2017, when I wrote the first version of this article, I decided that the term “Local AI”, which I had used for many years, was clumsy, and decided to draw inspiration from what happened in the world of computing, reflecting on the development of the Personal Computer. Hence , I called the concept of AI on the Edge “Personal AI”. Personal AI Not long after publishing the article , we immediately saw the advancement of Personal A.I. For many years prior to that, it was felt that CPU’s were becoming a commodity, and that the real innovation lied elsewhere. As an Engineering graduate who loved chips, I was dismayed by that. I still get excited when the latest desktop CPUs are announced and keenly look out for new developments in this space. Continued on page 41

2 Days 800+ Delegates 11 Keynotes 56 Talks 8 Workshops Demo Zone Exhibition Innovation Café


4-5 SEP2019


WELCOME After the amazing success of AI Expo Africa 2018, we are proud to announce our 2019 event will be staged again in Cape Town on the 4th and 5th September with a VIP opening event on the evening of 3rd September.

Why attend? The community feedback and media coverage from the inaugural event has cemented AI Expo Africa as "the" largest business focused AI gathering in Africa. AI Expo Africa 2019 is focusing on real world applications and trends driving the exponential Economy in Africa and is regionally focused. You will enjoy a packed two-day programme allied to a unique exhibition area aording delegates, exhibitors and sponsors learning and networking opportunities built around our 6 core themes, namely;

Innovation The Innovation CafĂŠ showcasing the African AI startup ecosystem

Knowledge Exchanging & sharing state of play of AI in Africa now & the future

Connecting Meet AI businesses, start ups, investors, deployment & service providers

Education How people and business can educate themselves about disruptive AI trends

Support The role Government, skills, training & investment will play in growing AI in Africa

Business Organisations seeking AI solutions with technology suppliers & platform vendors



Testimonials “Exceptional experience for LightBlue to position IBM Watson and Live Person - truly great opportunity to learn and exchange ideas”

“Great event, had a fantastic panel discussion regarding youth and woman” BRIGITTE BINNEMAN, TECHNOLOGY INNOVATION AGENCY


“Amazing conference Nick and Roy and team! Can’t wait to be back in 2019!” ZACHARIAH GEORGE, STARTUP BOOTCAMP AFRICA

“An awesome event. Great job by Nick and team” EKOW DUKER, IXIO ANALYTICS


“Such an awesome 2 days. Well done to you and all your team for an excellent conference. Venue also really good”

“Excellent conference, thanks Nick. Looking forward to next year”



Listen to Platinum Sponsor, Brett StClair, CEO SIATIK, describe the value they obtained from attending

Watch Clip - Click Here



Sponsors & Community Partners At our inaugural 2018 event we were supported by over 40+ sponsors and 20+ community partners. In our survey our community reported an extremely high level of satisfaction.



of respondents said they




of respondents said AI EXPO AFRICA MET ALL THEIR






rating the OVERALL


as Good or Excellent

rating the PROGRAMME MIX AND QUALITY OF TALKS as Good or Excellent







The Programme AI Expo 2019 builds upon the successful format from 2018 with the addition of an extra track dedicated to technology demos and the inclusion of executive AI master class workshops.

The 5 track programme for 2019:

We will be showcasing innovations and real world applications of the complete spectrum of AI / Cloud based technologies such as; automation, agents, NLP / NLG, interfaces, machine / deep learning, devices, voice interfaces, IoT, RPA, analytics, compute, cloud platforms, hardware, components and APIs.

Become a speaker We invite speakers from all corners of the AI community, including vendors, start ups, platform providers, researchers, policy makers, investors, Government, education, service providers and sponsors to submit papers for our packed programme.

The Venue

> > > > >


Submit your 150-word abstract and title for review today:

AI Expo Africa 2019 will be held at the prestigious Century City Conference Centre which combines the convenience of conference venue, hotel and social surroundings in the heart of the newest development in Cape Town:









Awards Kree Govendor

Lee Naik

Stefan Staffen

Google Cloud







Best Plenary Keynote Transunion

Best Track Keynote Microsoft

Best Flash Talk BCX

Best Workshop

Best Product Demo

Google Cloud

Data Prophet

Darlington Akogo

Philip Coetzee





Best Expo Stand

Best Innovation Café Stand

AI Expo Africa Director’s Commendation

Prize Draw Winner

Selected Press & Media

Contact us to discuss your package needs and pricing

Zindi and the Machine Intelligence Institute of Africa (MIIA) partner to transform Africa through Data Science Zindi ( and MIIA ( are pleased to announce a new partnership.


Zindi is the first data science competition platform for Africa. Launched in September 2018, Zindi hosts an online community of 1000 data scientists from across Africa and beyond. Currently there are five competitions open to the public with $14,000 USD in prize money up for grabs, with new prized competitions rolling out each month.



FOUNDED IN 2015, MIIA is a non-profit organization whose vision is to accelerate machine intelligence and data science research and applications to help transform Africa. MIIA aims to accelerate and deliver breakthrough research and practical applications to solve African problems, support entrepreneurial activities, and help drive long-term inclusive and sustainable scientific, technological and socio-economic development on the continent. Through this partnership, Zindi and MIIA will collaborate on designing new Zindi competitions which will launch in the coming months in technical areas such as deep learning using GIS data and satellite imagery. The two organizations plan to explore other ways to join forces to continue to strengthen the African data science ecosystem through in-person events and online resources. “MIIA and Zindi share a common vision for Africa in the era of big data. We are thrilled to work with MIIA to ensure

that data scientists in Africa have opportunities to hone their skills on real world problems, connect with their peers in other countries, build their professional profiles, and connect with valuable employment opportunities,” says Celina Lee, CEO of Zindi. “The support that MIIA and Zindi can provide to African businesses, non-profits and the public sector through the transformative power of data science, artificial intelligence, and smart technology can play an instrumental role in transforming Africa and help shape a better future for the continent in the Smart Technology Era,” says Dr Jacques Ludik, President and Founder of MIIA. To learn more visit the Zindi site ( or the MIIA site ( For a free consultation on hosting a Zindi competition, contact zindi@ ai

How to Address the Massive

Shortage of Data Science Skills

in Africa and the World: Just Do It

Distributed Ledgers and Artificial Intelligence led by Dr. Bitange Ndemo. To companies looking to leverage and develop AI talent on the continent: look at successful models such as InstaDeep, the trailblazing company led by Karim Beguir and Muthoni Wanyioke. They are sourcing and developing some of the best talent in Africa to solve data science projects globally. Also, Darlington Akogo’s minoHealth is an African AI startup doing great work in healthcare. To non-profit and governmental initiatives looking to develop AI talent: take a close look at Data Science Nigeria (DSN). The initiative exemplifies my thesis of how action through coding is the real gateway to an AI-driven society. DSN is a non-profit led my Dr. Bayo Adekanmbi. The initiative organizes machinelearning workshops and bootcamps in which participants compete on real-world industry problems via the Kaggle platform. I am privileged to serve on the advisory board of the DSN and I see it as a model for how Africa as a whole can take a global lead in AI talent development. Continued on page 28


Once one is done with the inspiring abstractions and swaths of eloquent poetry about how AI will transform our lives, one must arrive at the place of work. The place of implementation. The place where stuff really does happen. The place of action, of coding, of Keras, Tensforflow, SQL and noSQL databases, python, imageNet and transfer-learning. The place of GANs. The place of JAVA, JavaScript, Kotlin, Swift, and Computer Machine Language. The place of web services, RESTful APIs, Docker, and Kubernettes,. The place of CAD/CAM and CNC. The place of JSON, XML, protobuffs, etcetera, etcetera. It is at this place of coding work that the fourth industrial revolution will be realized or lost. It is at this very place of code that some nations will be catapulted into a future even more amazing than we could have imagined. It is also at this place of code that some nations may be left even further behind. This work of coding is therefore the gateway to an AIpowered future. And the gatekeepers are data scientists and data engineers. Let me be clear: No country on earth currently has enough data scientists and data engineers. Some have more than others. Africa is lagging behind in this area. However, in my view, Africa has the greatest potential to develop data science talent for itself and for the world. Of note, Africa is a populous and growing continent whose population is young and eager to learn. According to a 2012 U.N report, the median age in Africa was about 19.7 years old. The young people are eager for opportunities and simply need appropriate preparation and guidance. The secret sauce to building a skilled data science workforce is simple. Just do it. This is true on both a systems/government/ policy level as well as on an individual level. To the aspiring data scientist: don’t wait or expect to be taught what you need to know in some school or online certification program. Instead just start working on that project you have dreamt up. And if you have not dreamt up a project, you can start by training the MNIST dataset which recognizes hand-written digits. Google, Stackoverflow, GitHub, are all a dear friend of the aspiring data scientist. To the tertiary institution, government, or other policymaking body: guide and incentivize your charges towards the type of self-learning I described above. Also look for successful pioneering models, such as the Kenyan Taskforce on

By Dr Stephen Odaibo


Data science is of vital importance to the realization of the fourth industrial revolution. The term fourth industrial revolution was coined by the World Economic Forum, and refers to the artificial intelligencedriven melding of the biological computational spheres. Such an AI-powered future has enormous potential for good. For instance, through advanced Agricultural Technology AI can bring about food security making hunger a thing of the past. AI can also tremendously improve healthcare by automatically detecting diseases at their very onset. At the core of this revolution is data. Therefore the people with the computational skills to extract actionable intelligence from data are in increasingly high demand.

The Good, the Bad and the Ugly of Applying Ai in Customer Services Genii Ai, a Cape Town-based machine learning start-up has developed and successfully commercialised predictive models that inform customer behaviours to improve customer sales, services and retention. Customer expectations are rapidly evolving and demands for fast and convenient service have never been higher. This has a tremendous impact on Customer Services.

AT THE recent Ai Expo Africa 2018 Genii Ai shared the Business Case, Process Applied, Outcomes, Failures, Lessons Learned and Success Story of developing the first Customer Service Demand Prediction Model for a global Telecoms company. This project’s ROI was forecast to be >550% over the next 3 years, based on the results that were achieved during the first 3 months. Genii AI presented a real case study – NO THEORY – where Ai was applied to predict Customer Service demands across the Customer Journey and actioned to improve CX, and reduce customer effort. Ai and Machine Learning are nothing new, but applying Ai to solve business challenges, with a clear ROI, is quite unique in the customer services arena. Genii Ai has done just that and continues to be pioneers of applied Ai to solve real business problems in the 4th Industrial Revolution. They have managed to develop Ai solutions that not only de-risk clients during the deployment phase but are cost effective in the way they feed and process the algorithms and hence making Ai much more accessible to a greater market to leverage off the benefits of AI. ai To find out more about this exciting compnay and its AI offerings, visit:

How to Address the Massive Shortage of Data Science Skills in Africa and the World: Just Do It 4TH QUARTER 2018

Continued from page 27



Once the data science and data engineering talent base is built up to a strong level, the interface with other industries will occur naturally over time. Healthcare is one area where AI will bring about much needed transformation in Africa. I’ve been blessed to be a physician/retina specialist who is also a handson full-stack AI engineer. This combination of skills allowed my

company, RETINA-AI, to take a leading role in AI innovation in the eye-care space. We developed Fluid Intelligence, the world’s first mobile AI app for eye care providers. And we are now looking to employ talented people to join our growing team. It was this search that opened my eyes to the global skills shortage. There is a pressing need for all of us to join hands in building up the African data science skills bank. Our beautiful continent deserves it. The world deserves it.

BIO: Dr. Stephen G. Odaibo is Founder, CEO, and Chief Software Architect of RETINA-AI, a company using Artificial Intelligence to improve Healthcare. He is a Retina specialist, Mathematician, Computer Scientist, and Full-Stack AI Engineer. Dr. Odaibo is the only Ophthalmologist in the world with advanced degrees in both Mathematics and Computer Science. In 2017 UAB College of Arts and Sciences awarded Dr. Odaibo its highest honor, the Distinguished Alumni Achievement Award. In 2005 he won the Barrie Hurwitz Award for Excellence in Clinical Neurology at Duke Univ. School of Medicine where he topped the class in Neurology and in Pediatrics. In 2016 Dr. Odaibo delivered the Opening Keynote address at the Global Ophthalmologists Meeting in Osaka Japan. And he delivered the inaugural Special Guest Lecture in Ophthalmology at the University of Ilorin, Nigeria. In 2018, Dr. Odaibo delivered the keynote address at the National Medical Association’s New Innovations in Ophthalmology Session. He is author of the book “Quantum Mechanics and the MRI Machine’’ (2012), and of the book “The Form of Finite Groups: A Course on Finite Group Theory” (2016). Dr. Odaibo was born in Ilorin Nigeria, and currently serves on the advisory board of Data Science Nigeria.

How AI is Transforming Healthcare in Africa Q & A with Dr Jacques Ludik

Q: How is AI and machine learning transforming healthcare? Healthcare is a sector where AI has endless possibilities, from user-friendly bots and chatbots assisting patients in their health diagnosis to robots performing operations with precision. AI is currently used in the healthcare industry to identify high-risk patient groups, predict diseases, increase speed and accuracy of treatment and to automate diagnostic tests. AI has huge potential to improve drug formulations, predictive care, and DNA analysis that has the power to positively impact the quality of healthcare and affect human lives. Healthcare is also one of the key industries for Cortex Logic with productized solutions such as AI Risk Assessment for Hospital Benefit Management as well as Healthcare companies within Cortex Logic’s portfolio such as Mosaic, providing AI-based Precision Medicine for Oncology, SynerG, a Healthcare Data & Analytics Platform for Africa, and Foster, a AI-based psychosocially tuned therapeutic chatbot companion. Q: What are some of the prohibitive barriers to AI adoption in Africa? Are they the same as developed countries? Although Artificial Intelligence in Africa is on a roll and there is a realization amongst most South African business executives that they need to boost their organization’s competitiveness by innovating through investments in AI technologies, the reality in South Africa is that there is much planning, investment and policymaking required before local businesses and decision makers can properly leverage this transformative technology. For any South African business to stay relevant and thrive given the swift pace of change and disruption in the Smart Technology Era (4th Industrial Revolution), it needs to be transformed into a smart-data-driven business and have increasingly more real-time intelligence on all

Q: A shortage of skilled healthcare professionals is a problem facing just about every country. How do you think AI can help to alleviate that, especially in Africa? As smart technology such as AI makes it possible to provide more cost-efective services and scale with limited resources, it is strongly recommended that we combine healthcare and smart technology skills, knowledge and resources to develop and roll-out relevant healthcare related services for the African continent. It is an opportunity to leapfrog and provide impactful and cost-effective solutions that can also be rolled out internationally. AI can also be used to accelerate and improve higher quality education in healthcare and smart technology with more personalized and relevant content and guidance. South Africa also needs to fundamentally shift its thinking about AI and strategically plan to create a vibrant ecosystem where AI flourishes. Corporate South Africa will need to work with public institutions, educators and leaders to create a comprehensive long-term vision of the role of AI in the country’s economic development. Machine Intelligence Institute of Africa (MIIA) is an innovative community and accelerator for Machine Intelligence and Data Science Research and Applications to help transform Africa. MIIA aims to achieve this by networking together the critical mass of resources, promote and sponsor learning activities, and strengthen scientific and technological excellence, mentoring and collaboration on the continent. A core focus of MIIA’s activities is to accelerate and deliver breakthrough Machine Intelligence and Data Science research and practical applications to solve African problems, support entrepreneurial activity, and help drive longterm inclusive and sustainable scientific, technological and socioeconomic development on the continent. For maximum impact, MIIA is working with a range of partners where there is synergy with MIIA’s vision and mission and includes other nonprofit organizations, universities, research organizations, startup incubators, business and government organizations.


Cortex Logic is an AI software and solutions company that provides an AI Engine for Business to solves strategic and operationally relevant problems through operationalizing Data Science, Internet of Things (IoT) and Big Data & Analytics and delivering state-of-the-art AI-based applications, solutions and products. Cortex Logic helps platform businesses, corporates and organizations to thrive by operationalizing AI and unlocking the value from all available structured and unstructured data, all in support of its mission to help shape a better future in the Smart Technology Era.

aspects of its internal operations, customer needs and impact, and competitive and collaborative forces in the ecosystem in which the business operates. Businesses that are dissatisfied with their speed and/ or inability to unlock value from AI and solve strategic and operationally relevant problems, should collaborate with AI / Smart Technology partners such as Cortex Logic that can help them to accelerate the operationalization of AI solutions. Some of the problems that many businesses face include not having access to people with AI and data science related skills, struggling to attract top talent and cultivating the best possible environment for the top talent, having business divisions that are operating in silos, not having a solid, scalable data infrastructure and smart data layers to enable rapid access to all data for analytics, and not having the ability to integrate AI solutions into business workflows, processes and customer facing channels. Compared to the developed world, South African businesses in general also do not have high quality data streams. In addition to driving awareness of AI and skills training within the fields of data science and analytics, local businesses also have to ensure that they create and enforce strong policies around internal data storage and data protection.


Q: Tell us a bit about your background and role at Cortex Logic. Dr Jacques Ludik is a smart technology entrepreneur, Artificial Intelligence (AI) expert, investor and ecosystem builder with a Ph.D. in Computer Science (AI) and 25+ years’ experience in AI & Data Science and its applications. He is currently, amongst other roles, the Founder & CEO of Cortex Logic and Founder & President of the Machine Intelligence Institute of Africa (MIIA) and has also founded Bennit.AI, The Talent Index, Mosaic, SynerG and CSense Systems (Africa’s first AI company that was sold to a multi-national company, specifically General Electric in 2011). Apart from his executive management responsibilities in these companies, he was previously Vice President Data Science & Chief Data Officer at Jumo, Director and Big Data & Analytics Leader at General Electric and Senior Lecturer & Researcher at Stellenbosch University.


Q: Is health tech adoption largely being driven by consumers? If so, how? Yes, it seems like health tech adoption is largely being driven by consumers, especially in the wearable healthcare space. Although Smart technology related healthcare start-ups across the globe are attracting significant investment, there still seems to be a low AI adoption in the Healthcare sector in general when compared to other industries such as Telecommunication, High technology, Automotive and Financial services. Some of the characteristics of early AI adopters include companies that are digitally mature, typically larger businesses, adopting AI and other smart technologies in core activities, focusing on growth over strategies and having C-level support for AI. Some examples of value that can be created across the Healthcare value chain includes: • Predict disease, identify high-risk patient groups, and launch prevention therapies • Automate and optimize hospital operations; automate diagnostic tests and make them faster and more accurate • Predict cost more accurately, focus on patients’ risk reduction • Adapt therapies and drug formulations to patients, use virtual agents to help patients navigate their hospital journey



Q: These new technologies will require new skills outside of medical qualifications. What are the future skills that will shape the industry in the near future? Some of the new skills outside the medical qualifications does not only include technical skills such as those from data scientists, data translates, data engineers, AI and machine learning engineers, IoT specialists, but also psychology and skills related to emotional intelligence and communication. For students, a future-focused curriculum is a necessity. The World Economic Forum identified 16 skills that are needed in the 21st century—including creativity, collaboration, initiative, and adaptability—but are not included in standard curricula. People are also questioning the “learn then earn” model, wondering if lengthy degree programs still make sense in a world of fast-changing jobs. Instead, there are calls for a new deal on lifelong learning. The implication of these changes is clear: companies need to update the skills in their workforces, and individuals need to acquire skills that work with, not compete against, machines. For people already in the workforce, reskilling will be essential. Much of this reskilling can occur on the job through stronger professional development programs. For people transitioning between jobs, vocational and adult education programs must be strengthened. These work best when they are short, affordable, and closely linked to the job market. Nanodegree programs are one recent innovation in this space. Before the medical profession can realize the full potential of AI, health care providers must adopt significant changes in the way they do business, commit to a substantial investment in computing power and technical expertise, and work to increase the availability of the fuel that will power progress: data, including medical records. Q: After AI, what do you predict will be the next big health tech innovation? The next big health technology innovation will be a fusion of smart technologies that includes AI, biotech, neurotech, nanotech, virtual and augmented reality, blockchain, IoT, 3D printing, robotics, drones, quantum computing, etc. With all the investment and talent focusing on AI, one can also expect significant enhancements and breakthroughs in AI technology that would lead to more value adding applications. Some of the new exciting AI Applications in Biomedical Engineering includes: • Computational Pathology/Tumour detection: AI plays a major role

in lesion identification and classification, serving as a good prediagnosis tool in precision cancer care. Some of the tasks within this scope are: • Tumour Identification: This is a classification tasks that assigns the identified area into one of many output classes.It can also be a two-class problem where the task is to say if the tumour is present or not present (tumour detection). • Segmentation / Tracking: Identify areas of tumour or any other anatomy of interest in images of any modality - MRI, pathology slides, Ultrasound, etc. (U-Nets) • Registration - Multi-modal registration of images using Convolutional Neural Networks. Especially for non rigid registration • Touchless interaction in OR (Operating Room) - Machine Learning approaches are used to solve pattern recognition problems of capturing motions and gestures and using them to interact with modalities in the operating room. • Navigation in Image Guided Surgeries: Endoscopy is one example where machine learning approaches are used to identify the pose and scene and relate it to the point in surgical workflow. This can be used to predict surgery times and other aspects of the remaining surgical workflow. • Health-care robotics (Assisted surgery, haptic rehabilitation systems, laboratory automation systems) • Drug discovery • Protein Folding • Sequence Analysis (gene finding, multiple dataset integration) • Brain-Computer Interface and Neuroprosthetics • Experts Systems for advising health-care professionals • Epidemiological Data inference (e.g. tracking epidemics, finding patterns of exposure/symptoms) Q: What can delegates expect from your talk at the AI Africa Expo. As Cortex Logic is an AI Engine for Business, we are wellpoised to operationalize AI and help platform businesses and large enterprises thrive in the Smart Technology Era. AI Expo Africa is also a great opportunity to see how AI is now impacting many aspects of Commerce and Enterprise in Africa, and delegates will gain great value from attending. I’ll be sharing some practical applications of AI technology in various industries. • Get good perspective on operationalizing Big Data & Analytics, IoT and AI and its benefits • Discover how end-to-end AI solutions can unlock value from all available structured and unstructured data to: • increase operational efficiency, effectiveness and revenue • create strategic value via faster, better and more proactive decisions, enhanced scalability, new business models, and revenue growth opportunities • enhance customer experience via real-time, on demand, digital, personalized service delivery, assistance and advice which is enabled via 360 degree insights about the customer • enable more targeted sales and marketing • Get information on some of these solutions include strategic business transformation & optimization, human capital valuation & employee profiling, intelligent virtual assistants, robo-advisors, process optimization, predictive maintenance, fraud detection, churn prediction, advanced risk scoring, machine learning-based trading, real-time customer insights, smart recommendations and purchase prediction, personalized search, cyber security, medical risk prediction, and precision medicine. • Working collaboratively with innovate AI partners that can help business thrive in the Smart Technology Era and be agile, innovative and adapt quickly to stay relevant given the swift pace of change and disruption to business and society. ai Excerpts from this interview appeared as a radio interview on eHealth.

Pathways for the Adoption of AI by Managers in Africa A new way of producing goods and providing services using processes enabled by intelligent machines is raising both fear and hope among the managerial class of leaders of organisations on the African continent. This way is what is the current vernacular refers to as the Fourth industrial revolution (4IR).

vast availability of data, forces that are bringing about an era of machines that may match and in some instances exceed human intelligence. Examples of AI technologies are; computer vision, audio processing, natural-language processing, machine learning and expert systems.

The above are impacting traditional industries in areas such as: Self-driving cars Self- propelling rockets Biotechnology Genetic engineering Cryptocurrency Cybersecurity Computers that talk and think Global high speed internet Participatory Awareness Executives and other managers need to acquire a basic understanding of AI. Today’s manager needs to understand how programs learn from data to gain critical understanding of the basic architecture that completes a functioning AI system.


We are currently utilizing AI applications such as: Virtual Agents - Chatbots replacing call center agents. Cognitive Robotics that learn from their environment and interact with humans and other robots. Speech Analytics software that identifies emotions and stress. Identity Analytics - solutions that help define access to critical data and systems. Recommendation Systems for social media marketing and content targeting. Data Virtualisation - retrieval and manipulation of data by apps.


A NEW way of producing goods and providing services using processes enabled by intelligent machines is raising both fear and hope among the managerial class of leaders of organisations on the African continent. This way is what is the current vernacular refers to as the Fourth industrial revolution (4IR). 4IR represents a production system that is emerging out of advances in technologies such as; Artificial intelligence, Data Science, Connectivity and Mobility. These technologies are occasioning a new world where process capabilities between man and machine are becoming blurred into a similarity. Exactly what Alan Turing predicted in the 1950s. Turing proffered that if a human sustains a conversation between Computer A and Human B without being able to discern the difference between the two, then the machine would be said to have achieved intelligence of the order of humans. Heading into 2019, 70 years after the emergence of AI, the question on many people’s lips is: ‘Where is Africa’s place in this?”. The continent still suffers from the lingering effects of the 3rd Industrial Revolution; effects such as poverty, unemployment and environmental degradation. This dubious backdrop presents the African manager with a responsibility to draft and chart AI pathways that will lead to the social and economic advancement of the continent and its citizens. This article is explores some of the ways in which African managers can of ramp their organisations into the 4IR. There are several definitions of AI. A simple one is that: AI is a scientific understanding of the underlying mechanisms that lead to thought and intelligent behavior and their embodiment in machines. AI has emerged as a consequence of advancements in computing power, sophisticated algorithms and the

By Dr. Allen Mutono (PhD) Chief Intelligence Officer, CAI (Centre for Artificial Intelligence)

SAVE THE DATE AI Expo Africa 2019 4 & 5 September 2019 Century City Conference Centre, Cape Town

Organise for AI Organisations intending to adopt AI will also need to attract or develop teams of people with competencies in the various facets of AI. Ensure user Trust AI capabilities are similar to many digital initiatives that depend on data about people. Users need to be confident that organisations will respect and safeguard the information generated from the interactions. AI Health Check Checks need to be conducted across processes to enabling infrastructure and technical skills. As with many digital initiatives, success with AI depends on access to data sources both internal and external and investments in data infrastructure. Scenario Planning AI will shift the ways in which businesses generate value across markets, processes and functions. Organisations need to think even more expansively about their businesses, build cohesive future scenarios, and test the resilience of their plans against such scenarios.


Workforce Focus AI stands to create discomfort at work. The threat to jobs and careers in their current form is real and may lead to user rejection. Establishing an ‘all hands on the deck’ plan across teams encourages ownership and acceptability.



Re-Skilling teams We need to be re-skilling our people into new occupations. Managers need to look out for new competencies to train their team in. Conclusion There is no doubt that artificial intelligence will significantly impact the world. We don’t know where this is going, we don’t know what the social costs and consequences will be, we don’t know what will be the next area of life to be affected, the next business model to be destroyed, the next company to collapse or the next set of tools and techniques to become available to the people who seek to destroy others. What is certain is that all managers need to manage the future by developing viable and coherent AI plans. ai

REFERENCES: Abdul Razack, How Artificial Intelligence Will Transform Your Business, Business Insider Melbourne, 2016 Diana Bersohn and Mccree Lake, How To Build It Competencies For The Ai Era CIO Notes Johannesburg. 2017 Enrique Suarez, 7 Uses Of Exponential Technology We’re Excited To Watch In 2017, Singularity University San Francisco, 2017 Jacques Bughin, Artificial Intelligence The Next Digital Frontier? Mckinsey Global Institute Brussels, 2017 James Scott, Artificial Intelligence – How Can It Benefit Africa? Businessbrief Johannesburg, 2017 Mark Purdy and Paul Daugherty, Why Artificial Intelligence Is The Future Of Growth Accenture, Johannesburg 2017 Nick Bostrom, Superintelligence: Paths, Dangers, Strategies, Oxford University Press, Oxford 2014 Nouha Abardazzou, The Rise Of Artificial Intelligence In Africa, How We Made It In Africa, Cape Town, 2017 Ray Kurzweil, The Singularity Is Near: When Humans Transcend Biology Penguin Publishing Group, 2005 Sam Ransbotham, David Kiron, Phillip Gerbert And Martin Reeves, Reshaping Business With Artificial Intelligence MITsloan Management Review Cambridge, 2017 Stephen Time, 6 Artificial Intelligence Startups In Africa To Look Out For, Digital All Stars Cape Town, 2017 Willie Schoeman, Artificial Intelligence Is South Africa Ready, Accenture/ Gordon Institute of Business Johannesburg, 2017

The Potential for AI in Africa

Health-Tech For AI enthusiasts, and experts, this shift opens opportunities to improve human welfare and quality of life. AI in Africa can help us diagnose diseases in rural

villages, and connect low-income earners to a health grid that was impossible years ago. It will deliver nursing care for the vulnerable, elderly, and help manage local infrastructure in rural communities with instant results and measurable insights. The key to achieving this goal in the health care space is trained applications and solutions that focus on data and instant accessible research information. Before AI systems can be deployed in healthcare applications, they need to be ‘trained’ through data that is generated from patients in rural villages Example: I was speaking to a startup in Africa recently that has developed a business case and started providing instant diagnosis of a number of diseases for rural communities in certain countries and urging the patients to enter that data via SMS or USSD which is then processed and sent to the AI back-end and processed in real-time to provide first live diagnosis and prescription for the patient before even seeing a doctor. The project is currently in beta stage but from an emerging market perspective these are the kind of solutions that excite me about what AI brings to our world. Within hospitals a number of activities, such as screening, diagnosis, treatment assignment and so on can be programmed using AI to achieve better results and diagnosis but also reduce a number of deaths based on mis-diagnosis. The patient data often exists, but access to the data as well as the time to diagnosis is one of the issues AI can help us solve as well as bridge the gap in illness discovering. Digitization of health-related records is accelerating from the sharing of high quality, labeled and specialized data sets to development of deep learning that is highly adaptable for integrative analysis of heterogeneous data sets created from diverse sources.  Deep learning itself is being used to rapidly combine


THE OBJECTIVE to create an inclusive society especially in emerging markets is where I see the major value of Artificial Intelligence. Major technology companies are investing heavily, and competing to bring AI into our lives. They are setting up research labs around the world, and snatching up some of the best talents to be trained on AI and a variety of use-cases. Companies like Apple, Amazon and also a number companies from leading markets in Asia i.e Xinhua News Agency, Alibaba, and Wechat. With recent breakthroughs in machine learning, and robotic language communication we are seeing some of the ways in which organizations are leveraging big data, and we are seeing a shift in how we do things in our everyday lives. In emerging markets like Africa we are seeing AI being deployed in Health-tech, Fintech and a number of use case in the Agri-tech space as well.

by John Kamara City AI Ambassador: Lagos and Nairobi, Founder : Afrimart, Jamborow.


Human behavior in the next 50 years will be defined by robotic personalities based on artificial intelligence. AI will guide our homes, cities, life-styles and other human experiences. It will offer us more satisfying jobs, and free us from monotonous tasks thereby allowing us to focus on the creative and social tasks that AI is incapable of. This is based on a discussion I had recently with a group of friends but my question still remains. Is this assumption true?

legacy information and a number of different programs to develop cures for a number of diseases and run quicker objective test trial simulation without the use of human subjects.


Fintech The fintech space in Africa is disrupting everything we know, and traditional institutions are running around trying to play catch up. This new industrial revolution is highly disruptive, spawning social, and financial transformation of over 800m people in Africa who live on the low-income belt and aren’t financial included because they could not be qualified in the traditional ways in the past. We are seeing in Asia (China, India, Malaysia etc) the growth of the fintech space driven by AI solutions around micro-lending and other nano products for the lowincome earners in real-time. The new form of evaluation based on a number of data sets that we could not think



of using to create credit ratings for an individual 10 years ago is also setting a trend for how AI is helping us reshape this new financial age. Emerging markets like Africa have the raw materials needed to fuel the growth of Fintech (critical mass of people), and the use case required to create an AI-led financial engine that will control the credit value of the continent over the next 10 years. This also applies across emerging SME markets; SMEs in Africa are debunking the myth that it is risky to extend credit to them based on AI-led platforms that have blossomed in countries like Nigeria, Kenya, Ghana, Tanzania. Other countries like Liberia, Sierria-Leon, Ethiopia, Rwanda etc are also following suit as well. Example: One of the success stories in Kenya is Mpesa but we also have M-Shwari a partnership between Commercial Bank of Africa and Safaricom which offers digital banking and short-term credit services tailored to the poor in

Kenya. This was one of the first scalable micro-lending (fintech-based) solutions. In just 5 short years, the service has grown to 20 million customers and almost a billion dollars in micro-loans disbursed, supported by automated credit assessments and instantaneous approval decisions. Using AI-led risk measure, and data defined by the individual’s life and environmental value, we are now seeing a new type of economy developing in a number of African countries. It is estimated that this fintech led economy could be worth over $250bn by 2021 Companies like Branch, Tala, Paylater, Jamborow etc have also come into the market to provide a number of alternative solutions using AI as well. With better predictive models tapping into the alternative data that does exist, AI is helping small business owners circumvent the traditional data that does not. In the end, I do believe we will arrive at a future that

converges and aligns a world of humans, and machines based on levels of intelligence and ability to handle fundamental task. The argument around Ethical AI is a still a very valid one which must be thought through as we build this future based on AI. Machines will respond to the data set they are fed but will also become smarter to know that the data set does not give accuracy of value. This is where the question I asked in the beginning becomes interesting. ai

The real impact of AI on jobs Pessimistic forecasts predict that AI will cause massive job losses, doing for cerebral work what mechanisation did for manual work. Does actual data support such a grim outlook, or is it just alarmist hype?

By Andrew Quixley: Watson Sales Leader - IBM Middle East & Africa. Tech Advocate, Speaker & Writer Specialising in AI

‘ROBOTS ARE Coming For Your Job’ read the headline from an online article published last month. It’s just one of many alarming predictions that take a similar line — stoking fears about AI causing huge job losses — but this one has prompted me to discover whether actual data supports the pessimism, by examining the impact of employment disruptions in our past. Part 1 explores the impact of mechanisation on employment in manual labour.

Machinery, Tractors per 100 sq. km of arable land (green line, left axis) versus Employment in Agriculture, % of total employment (red line, right axis) in Spain from 1969–2009. The transition is very clear; a 558% increase in farm machinery, while employment in agriculture declined by 86%. Those changes took place in just 40 years — a span of less than one working lifetime. Similar patterns can be seen in the data for many other countries. Extrapolating to a worldwide scale, the data suggests that hundreds of millionsof farming jobs have been lost globally in the past two centuries. (Although there were many other innovations in agriculture, their impact was mostly on crop yields, rather than employment. Mechanisation had the biggest impact on employment). The key question is: did all this job displacement actually lead to commensurate increases in unemployment, or to widespread economichardship? Impact of mechanisation on unemployment The charts below are for six countries that experienced radical transformation in agriculture in the last 25 years — Brazil, China, Indonesia, Poland, Thailand and Turkey — comparing Unemployment, % of total labour force (blue lines) versus Employment in Agriculture, % of total employment (red lines) between 1991–2016.


Image licensed from


Part 1 — Mechanisation of manual labour The most profound change of employment patterns in the modern era was triggered by the first Industrial Revolution in the 18th and 19th centuries, when horse power gave way to steam power in Europe (and much later to the internal combustion engine). Steam-powered engines made mechanisation and productivity possible on an industrial scale, driving seismic transformation in employment patterns and lifestyles. Mechanisation certainly did displace manual jobs and the greatest impact was in agriculture. In pre-industrialised societies, up to three quarters of the workforce were (or are) involved in agriculture. Post-industrialisation, the figure is far lower. For example, in USA in 2016, according to the International Labour Organization (ILO), employment in agriculture as a percentage of all employment was just 1.62% — about one person for every 60 in employment. Historical data is not available for the countries that went through their industrial revolutions in the 18th and 19th centuries, but there is reliable data for those that made the transition much later. The chart below shows Agricultural



In all six examples there is a significant decline for employment in agriculture,and in four of the six cases there is also a downward trend for overall unemployment. We can surmise that all the displaced farm-workers in these countries found employment in other fields(!) assuming they didn’t all retire. In the two cases where there is a slight upward trend in unemployment, the downward trend for agriculture is far steeper. So for example, for every 23 people in Turkey who left employment in agriculture, only two show up in a rise in unemployment. Perhaps we shouldn’t be surprised that not every country was able to generate more jobs than were displaced; economies are the products of very many interrelated factors, and mechanisation in agriculture caused the greatest labour displacement in history. There are still deep scars in certain places. The oversupply of farm labour in USA in the 1920s was ultimately ascribed to several factors other

than mechanisation — according to the US Bureau of Labor Statistics in 1931 — but the effect of mechanisation surely exacerbated the situation. Notwithstanding the results in Indonesia and Turkey, that so many other countries have mechanised agriculture, displacing millions of jobs without triggering mass (or any) unemployment or recession, suggests that lessons have been learned and disaster is not inevitable. According to ILO, global unemployment stands at 5.7% in 2017. Impact of mechanisation on economies So how did these countries fare in terms of economic growth, while their agriculture sectors were shedding millions of jobs? The short answer is well. Without exception, all seven countries experienced vigorous economic growth over the 25-year period, weathering the global recession of 2008 (visible in the lines) and creating the right conditions for huge job creation.


Part 2 — Impact of automation ‘Powerful computers are now performing mental operations that could not possibly be accomplished by human minds. Any worker who now performs his task by following specific instructions can, in principle, be replaced by a machine’. The words could have been written last week about the current trend for AI-enabled automation. In fact, they were written before half of all the humans living today had been born. They are the words of Wassily Leontief — the Nobel Prize-winning economist — whose famous paper of 1983 anticipated computingenabled automation of routine work.

Unemployment in the OECD members was lower in 2016 than in 1991, and — while it shows greater variance than Least Developed Countries and the World — the long term trend line is approximately flat at 7%. The impact of the global recession in 2008 is clearly visible, affecting OECD members much more than Least Developed Countries, but there is no clear evidence of rising unemployment coinciding with the uptake of automation. The impact on the economies of these same groups is shown in the chart below. By 2016, GDP per capita for the OECD members had grown to 287% of its 1991 level and for the Least Developed Countries it was 301%, albeit off a much lower base. Again, there is no clear indication in the data that higher levels of automation were any hindrance to economic growth. The charts above are congruent with the conclusion reached by The Mannheim Centre for European Economic Research (ZEW) and the University of Utrecht, who assessed the impact of routine-replacing technological change (RRTC) on labour demand across 27 European countries between 1999 and 2010. According to their study — Racing With or Against the Machine? Evidence from Europe, published in 2016 — automation resulted in lower


Creation of jobs Tractors don’t just fall from the sky (except in tornado movies). Farm machinery has to be designed, manufactured, distributed, sold, delivered, fueled, tooled, serviced, repaired and replaced. New supply chains are needed — reaching all the way backto iron ore and crude oil — and these supply chains create jobs. And then there are other innovations into tangential domains — agronomy, animal feeds, consulting, crop research, fertiliser manufacture — that would have been impossible if three quarters of the workforce were still outdoors labouring in all weathers. As for indirect job creation, liberating so many of our forebears from toiling in the fields allowed much more time for thinking, innovation and invention, which all drove higher productivity and economic growth. Any employment vacuum was evidently filled by other kinds of employment, and we know this certainly happened because there are more than 3.2 billion people working today, compared with just 1 billion people estimated to be living in 1800. Mechanisation also triggered many other significant social changes. Working hours were far shorter — allowing time for a new pursuit called leisure, life expectancy was longer and the population grew rapidly from the middle of the 20th century. It also drove urbanisation and the expansion of cities. On this evidence, it’s safe to conclude that mechanisation in agriculture wasn’t good for humanity overall, it was absolutely great for humanity. If AI follows a similar path, we will have nothing to fear except alarmist hype that is trying to derail our progress. Without the progress that was only made possible by mechanisation, we would not be living in the world as we know it today, and the chances are, you would still be a farm-worker and probably not reading this, nor would I be writing it. In fact, it’s even more likely that neither you nor I would have been born!

The impact of automation so far Within a decade of Leontief’s paper, back office automation was well underway, automating transactional tasks in corporate finance, human resources and procurement functions. If the nineties were about back-office automation, the noughties were about front-office automation, tackling sales, marketing and customer service, and giving impetus to the global IT industry, which will see $3.5 trillion of IT spending in 2017, according to Gartner. So how did the trend for automation impact employment and the economy as a whole? If automation was a cause of unemployment, we would expect unemployment to have risen in countries which adopted higher levels of automation. Taking the OECD members as a proxy for most automated countries and the Least Developed Countries as a proxy for least automated countries, we can compare how unemployment has moved over the last 25 years in the chart below.


production costs and lower product prices, which increased the demand for products. The higher product demand in turn led to an increase in labour demand; indeed the paper reports that ‘RRTC is estimated to have raised labor demand by up to 11.6 million jobs across Europe, corresponding to almost half of the total observed employment increase over the 1999–2010 period’. Dr. Terry Gregory, senior researcher at ZEW and co-author of the study, said: “..labour has been racing with rather than against the machine.” Similar results have been observed in USA for robotics, another major driver of automation. The automotive sector is the largest consumer of industrial robots, at 35% of the total stock in 2016. According to the International Federation of Robotics, between 2010–2015 the USA’s automotive industry added 80,000 industrial robots. Over the same time period, the number of people employed in the automotive sector in USA increased by around 230,000, according to the US Bureau of Labor Statistics. On this evidence it appears that the major causes of automation over the last three decades — enterprise software and robotics — have had a positive impact on jobs and value creation. Can we extrapolate forward to predict similar results for AI?



Future impact of automation Contrary to the popular scare-mongering of the moment, ‘automation and digitalisation are unlikely to destroy large numbers of jobs’, according to an OECD report — The Risk of Automation for Jobs in OECD Countries* — published in 2016. The paper estimates that, on average across 21 OECD countries, 9% of jobs are ‘automatable’, but argues that this figure must ‘not be equated to actual or expected employment losses from technological advances’, for reasons detailed in the paper. The reasons in brief are: slow adoption, job-switching and job creation — phenomena that were also present during mass mechanisation. McKinsey Global Institute (MGI) takes an even more positive line, asserting that ‘automation could be the shot in the arm that the global economy sorely needs in the decades ahead’, in a study published in the Harvard Business Review in April 2017. The report goes on to estimate that ‘automation could increase global GDP growth by 0.8% to 1.4% annually’, and mitigate the effects of dwindling birth rates in developed countries, by adding workforce capacity even as populations decline.

Predictions for AI AI could be on track to contribute trillions of dollars to the global economy. The view is based on extrapolating to a worldwide scale the report commissioned by the UK government and published last month, which estimates that AI can contribute £630 billion ($837 billion) to the UK economy by 2035. That’s roughly £10k for every person living in UK today. AI can also improve employability for hundreds of millions of people, through its impact on education and skills development. The International Labor Organization estimates that 1.4 billion people are in vulnerable employment today, while UNESCO estimates that more than 600 million children and adolescents worldwide are not achieving minimum proficiency levels in reading and mathematics, and if the world is to achieve Sustainable Development Goal 4 — universal primary and secondary education by 2030 — we will need an extra 69 million teachers. It’s very hard to imagine the world conjuring 69 million human teachers by 2030, but we already have the capability to implement AI-powered teaching bots that can be scaled to provide personalised learning and skills development to everyone who needs it, liberating huge opportunity for communities that have been marginalised up to now. This has all happened before The studies cited here indicate that outcomes for technologyenabled automation in the modern age mirror the earlier outcomes driven by mechanisation of agriculture; mass unemployment and economic catastrophe were not among them. We humans have already survived a workplace transformation in which more than half of all jobs were extinguished, without causing mass unemployment or economic collapse, and emerged from that process much more prosperous, healthier and with more leisure time. Is it safe to base future predictions on past experience? Kenneth Rogoff, a former Chief Economist of the IMF and Professor of Economics and Public Policy at Harvard University wrote in 2012 ‘ should be careful in extrapolating the experience of the last two centuries to the next two’ but also goes on to say ‘..nothing suggests a massive upward shift in unemployment over the next few decades’. AI is just another technology, one more iteration in a long line of innovations and not the final station on the line. Is there any reason to assume that AI will be very different — in terms of its impact on economic growth and employment — to all technologies that have gone before? I don’t believe so. This has all happened before, and it turned out well for us. It will again. The opinions expressed here are my own. For all the charts, the data source was https://data. Full citation: Arntz, M., T. Gregory and U. Zierahn (2016), “The Risk of Automation for Jobs in OECD Countries: A Comparative Analysis”, OECD Social, Employment and Migration Working Papers, №189, OECD Publishing, Paris.

Ethical considerations for Artificial Intelligence There is little doubt that the pace of innovation is accelerating at unprecedented levels. Technology enabled breakthroughs are happening with increased frequency, enhancing the human life span, improving access to basic needs and leaving the general public with little time to adjust and comprehend the magnitude of these advances.

Data perpetuating embedded bias AI systems feed off data. If AI is the new electricity, data is the grid it runs on. AI systems look at data, evaluate that data against its goals and then find the most optimal path towards achieving those goals. Data is absolutely critical for AI systems to be effective. Machine learning (ML) algorithms gain their experience from the data they are given and if that data is biased or ethically or morally tainted, the ML algorithms will perpetuate this. What about factors that are not expressed with data, such as the value of another person, the value of connections, the value of a relationship? The biggest challenge with data unfortunately, is that data quite simply just does not give you ethics. Then there’s the issue of blame, who is to blame for AI making mistakes? The manufacturer, the software supplier, the reseller, the data set, the owner, the user? The issue gets more complicated when we talk about the loss of life in an accident. Consider incidents with AI systems in Healthcare and who would be legally held liable. What about autonomous vehicles disrupting the automotive industry and making its way into society sooner rather than later? If we expand on this trend, what about AI systems making decisions based on their programming, leading to them committing crimes? Are they guilty? Can an AI system be guilty of a crime? Are their programmers to blame? Their data sets? What laws govern this eventuality? Take our smartphone’s autocorrect function as a simple example. I’m positive many of us have had an incident where we’ve sent texts to friends right after an autocorrect function changes one word to another, often to a more embarrassing version, from where we often issue some grovelling apology. The point is this; if technology today struggles with understanding the intent of a few lines of text, how can we count on it to understand and make life and death decisions? Continued on page 40


Ethical fundamentals of everyday life The question of ethics finds some of its roots in the notion of fairness. What is fairness? How does one define fairness? Instinctively, human beings grasp the concept of what is fair and what is not fair. As an example, we commonly accept that one for me, two for you, is not fair. We teach our children about what it means to be fair, why we need to share, what the moral and ethical constructs as we believe them to be are when it comes to fairness and sharing. But how do we teach AI systems about fairness in the same way we teach our children about fairness, especially when an AI system decides that achieving its goal in an optimal manner can be done through unfair advantage? Consider an AI system in charge of ambulance response with the goal of servicing as many patients as possible. It’s quite possible that it might prioritise serving 10 people with small scratches and surface cuts above serving 2 people with severe internal injuries, because serving 10 people allows it to achieve its goal better. Although this optimises patient service, it fundamentally falls flat, when one considers the intent of what was meant to be accomplished in the most optimal way. In business we have ethical and unethical behaviour and we have strict codes of conduct regarding what we consider to be ethical and unethical business conduct. We accept that not everything that is legal is ethical and not everything that unethical is illegal and as a society we frown upon unethical business conduct, especially from big corporates. How does this transfer to AI systems? Surely,

we wouldn’t want AI systems that stay within the bounds of the law, but push as hard as they can against those boundaries to see what they can get away with, exploiting loopholes to fulfil their goals.


WITHIN THE field of Artificial Intelligence (AI), this phenomenon is certainly just as true, with the accelerated pace of AI development, generating huge interest about moral AI and how, as imperfect human beings, we are teaching AI the differences between right and wrong. As AI systems continue to evolve, humanity will place increasing levels of trust in them for decision making, especially as these systems transition from being perceived as mere tools, to operating as autonomous agents making autonomous decisions. The question of the ethics pertaining to the decisions that get made by AI systems must be addressed.

by Rudeon Snell, Leonardo Leader at SAP Africa

Intel Movidius Giveaway AI Expo 2018 TEBOGO NAKAMPE from ABI met with Gianluca Truda, a UCT Computer Science Post-Graduate, shortly after AI Expo 2018, to hand over an Intel Movidius neural compute stick. Gianluca attended the Intel Movidius NCS workshop at AI Expo Africa 2018, and was the lucky recipient of this great prize. Gianluca is interested in contributing to Intel AI Academy and the AI community in general, and he is currently working on a Computer Vision software application that he will share on Intel Devmesh once he is finished. Tebogo Nakampe (left) is seen here handing over the Intel Movidius NCS to Gianluca (right) to help him accelerate his CV software applications. ai

Ethical considerations for Artificial Intelligence


Continued from page 39



Revisiting classic questions regarding ethics and morality Researchers have explored how to effectively resolve this situation in the past. The Trolley problem tests have been around since 1967. First proposed by the philosopher Phillipa Foot it has subsequently proliferated into many variants. Generally, it is used to assess what actions people would take when asked to take an action that would, for example kill 1 person vs 10 people. This is specifically being applied in the context of autonomous vehicles, as a reference model to help AIs make effective life or death decisions, but it’s not a fool proof solution. Utilitarian principles could offer a framework to help with the ethical decisions AIs need to make. The focus would be on AIs making decisions that result in the greatest good for the greatest amount of people. However, at what cost? How do utilitarian calculations that violate individual rights get reconciled? Ethics is often not about one thing or the other specifically, but more leaning towards the notion of how if you go down a particular road, that road has a

particular set of ramifications. If you go down an alternative road the implications could be different. This is what AIs currently struggle with and what humans instinctively understand. AI systems have largely been built to achieve specific goals and specific outcomes. For humans to have any semblance of creating ethical AI, AI systems should be programmed to be sensitive to achieving its goals in the construct of human values as they could achieve their goals in rather bizarre fashions. Think about a machine deciding to protect humanity by enslaving it (The movie i-Robot rings a bell). Soft governance, industry standards, professional codes of conduct and policies. These are the considerations that must be given in order for us to understand how we can engineer AI in a safer way and how we make our values part of the design process when implementing AI systems. Who decides how ethics are defined? Who decides which ethics are applied in AI? Ethics ultimately is embodied in knowing the difference between what you have the right to do and what is right to do. We all will need to do our part in ensuring AI systems know how to do this. ai

Winner of DJI Mavic Air Drone announced The Lucky Winner of the DJI Mavic Air Drone at AI Expo Africa 2018 has been announced. Philip

Coetzee of Woolworths

received his prize from AI Media’s Nick Bradshaw, after having been selected in the Lucky Draw of Delegates who completed the AI Expo Africa Survey. Congratulations, Philip Safe Flying.

In 2017 however, we saw a surge of development in the mobile phone space where hardware was being included that could accelerate neural-network performance. Apple launched the iPhoneX that year with their A11 Bionic chip , which included a neural engine, and this was to facilitate the running of face detection algorithms in order to unlock phones. Many didn’t see this coming from a hardware perspective, however , if you were following the development of AI and guessed that Personal AI would happen, you would have known that this was merely the first step to greater things. Other manufacturers joined in , and hardware giant Intel announced a chip that simulates a neural network (Neuromorphic Computing), called Loihi. In parallel to this , the proliferation of AI digital assistants really took off. Simple devices like internet connected smart speakers started selling like crazy, and the speech recognition with internet search capability made it seem like you really

had a Star Trek ships computer at your disposal. Think of how normal it is today , to talk to one of these smart assistants ? How quickly did it become normal from only a few years ago ? Now, the continued development of the interface ( the smart assistant ) , using Cognitive Services like speech recognition , is an important parallel development to the hardware advancement on portable devices, in order to give us Personal A.I. What really got me excited, however, was that in early 2018, Gartner published an article saying that Artificial Intelligence was a game change for Personal Devices, a real vindication of what I was saying for years. According to the report “by 2022, personal devices will know more about an individual’s emotional state than his or her own family”. Not only was this the confirmation of what was happening, the use cases were starting to develop. There are so many uses for increased intelligence on the Edge that it will soon be taken for granted. By 2020, I see most pocket computers (what you may still call mobile phones) having advanced ML capability as standard to perform a whole host of tasks. The entry point into the capability will be the personal assistant, which will rapidly become much smarter, finally utilising the hardware for acceleration. Speech recognition will improve as it will be done on device. Instead of the personal assistant being just a front to a search engine in the cloud (as it is mostly today), the Personal AI will have real capability to process and understand, using the search engine to reference data. Continued on page 42


Continued from page 16


Personal A.I.



Personal A.I.

Best Product Demo

Continued from page 41 My original article predicted the following scenarios, and already they don’t seem so far fetched : • Realtime Health Analysis – currently your smartwatch monitors your heart rate and steps and sends it to your phone, which sends it into the cloud. In the near future, your Personal AI will read this data and analyze it in real-time, with the ability to alert you as early as possible should you be at risk of a heart attack , or stroke ,for example. Now I said, in 2017, that the number of complex sensors built into your smartwatch will increase in order to enable this, and lo and behold, in 2018 the Apple Watch 4 included capability to read ECG / EKG. Advancement in sensor technology will be key to deliver on the capability of Artificial Intelligence. Remember that the only reason we have the AI boom now is

discuss your s and pricing

because progress was made in the area of big data, and data is needed to train models. • Environmental Analysis – Another scenario that’s easy to predict will be to use the capability on your pocket computer to perform analysis of the environment around you. Air quality is a problem in many parts of the world – imagine taking out your device, some basic readings being taken, and then machine learning kicking in to advise you if the air is safe to breathe in that location, and what the risks are. This is especially useful for travelers. Once again, I am saying that development of better sensors is critical. Imagine a form of sensor that could analyze water in a glass and tell you if it is safe to drink – very useful in certain countries. • Realtime Language Translation – This is already happening with products like Skype, but could be augmented with the power available on the Edge device. I would imagine that a future version of Skype could take advantage of Personal AI, to improve Realtime translation (the already translated language is sent to the cloud and all the way to the other end). • Custom Apps – Once you have the processing power (and the sensors) at the Edge, you will see all kinds of custom apps being built to take advantage of this. Environmental Sensors could be utilized to create an app for workers in dangerous environments, mines for example. Apart from just the raw sensor readings, it is the Personal AI engine that would add real value, delivering insight in near real-time. It would also assist in sending more relevant data into a bigger engine in the cloud, with the knock-on effect of helping train more accurate models. The development of Personal A.I is an area of technological advancement that I believe will have more of a personal effect on your life. Just like the smartphone and social media did a decade back , the ability to both interact with a smarter personal assistant, and also get access to life changing services as listed above, will change our daily life. Perhaps a more natural interaction with technology will even stop us all from staring at a screen all day - we can only hope. ai




R 2018

ZAR25 | US$2 .50 | Euro1 .60

The Voice of African AI & Data Science







Advertise your Company and Services in Synapse Magazine







3RD QUARTER 2018 ZAR25 | US$2.50 | Euro1.60

The Voice of African AI & Data Science

Africa Africa '' ss Ar Ar tt ii ff ii cial cial Intelligence Intelligence Ma Ma g ga az z ii n n ee


AI has the Potential to Transform Africa, leveraging its benefits for the Social & Economic Good of the Continent. AI Expo Africa will see a gathering of Africa’s AI Leaders to plot the way forward for the 4th Industrial Revolution

Published Quarterly


Official Publication of AI Expo Africa Endorsed by the Machine Intelligence Institute of Africa






AD SIZE                             RATE

SELECT SELECT AD AD  SIZE SIZE (Mark (Mark with with X) X)

COMPANY LISTING (Logo, (Logo, Co. Co. Description Description & & Hyperlink) Hyperlink)

R2500                ___________

1/4 PAGE

R3500               ____________

1/2 1/2 PAGE PAGE

R4500        R4500                    ____________ ____________


R6500               ____________


R10 R10 000 000                     ____________ ____________

Company:  Company:

_______________________________________________________________________________________________ _______________________________________________________________________________________________


________________________________________________________________________________________________ ________________________________________________________________________________________________

VAT  VAT No.: No.:

______________________________________________ ______________________________________________


_____________________________________ _____________________________________

Person Responsible for Account: ____________________________________________ ____________________________________________ Print Print Name Name & & Sign Sign

____________________________________________ ____________________________________________

Date: Date:

______________________________ ______________________________

Advertising Advertising Artwork Artwork to to be be  Supplied Supplied as as High-Resolution High-Resolution PRESS-Ready PRESS-Ready pdf pdf -- At At least least 300dpi 300dpi Art & Editorial Features to be Submitted to

Ad & Editorial Enquiries -

Profile for AI Media Group

Synapse Issue 2 - 4th Quarter 2018