AIIMS A new generation of African talent
VIRTUAL HELPER
Bridging the language gap
AFRICAN UNION
Collaboration key for successful AI in Africa
GC REAIM
South Africa shaping responsible AI in the military
Announces inaugural AI for Good Impact Africa Event 2025 to be hosted in Johannesburg
Natasha Ochiel (Kenya) Co-Founder | CEO
The AI Centre of Excellence natasha@aiceafrica.com natasha-ochiel
In a bid to champion Artificial Intelligence in Africa, she co-founded the AI Centre of Excellence that aims to build sustainable value for AI in Africa by building capacity, building demand and building AI solutions through the Centre in Kenya.
Darlington Akogo (Ghana) CEO, minoHealth AI / KaraAgro AI & Member of United Nations ITU & WHO Focus Group on AI For Health darlingtonakogo
Darlington Akogo is a global leader in Artificial Intelligence. He's the Founder and Director of Artificial Intelligence at GUDRA, and its subsidiaries; minoHealth; an AI Healthtech company based in Ghana.
Naomi Molefe MSc, Manager: Strategic Sourcing and Research at Discovery & Co-Founder & Chapter Director Women In Big Data, South Africa
naomimolefe
N
aomi is the co-Founder and Chapter Director for women in Big Data SA; a registered NPO that is part of a global community of 17 000 women. The learning community works with strategic partners to cultivate tangible opportunities for women, unlock latent potential through accessible training and act as a catalyst for the advancement of women in Big Data fields.
INTERNATIONAL EDITOR
Deon Van Zyl (Norway) Senior System Developer Nordic Semiconductor
deonvanzyl
For 25 years, I have been an accomplished IT professional who is skilled in multiple languages with a strong background in development, security, digital forensics, automation, AI, management, and teaching. Through my work, I have been exposed to various industries, influenced public opinion, and networked with innovative leaders. As a roving Guest Editor, I write about new technologies that are reshaping the world such as Artificial Intelligence, VR & AR, Cybersecurity, Web3 and more.
Dr Nick Bradshaw Editor & CEO
AI Media Group
nickbradshaw
Nick is a Tech focused Executive / Entrepreneur helping clients, communities, practitioners & start-ups understand the value of Artificial Intelligence, Automation & Digital solutions in EMEA region. With 25 years experience in Europe, North America & Africa Nick has worked with a diverse set of Multi $bn global clients seeking to deploy and mature Enterprise grade software & cloud solutions. He is founder of the AI Media Group a new hybrid media, events, consulting & trade community offering clients insights into the growing 4IR opportunity in Africa. AI Media publish Synapse Magazine and run Africa’s largest Enterprise AI Trade Show – AI Expo Africa."
ABOUT SYNAPSE MAGAZINE
Synapse Magazine chronicles the 4th Industrial Revolution as it unfolds in Africa and plays a vital part in connecting the members of this rapidly growing trade community across the region. With a global readership, it puts Africa centre stage with a clear focus on the African 4IR innovation & investment narrative. We cover a range of technologies including; artificial intelligence (AI), intelligent automation (IA), robotic process automation (RPA), internet of things (IoT), big data, analytics & devices, as well as emerging standards, ethics and privacy concerns. Now entering its 6th year of publication, this unique quarterly trade publication is FREE to read on the ISSUU platform.
e are going to need a BIGGER boat!”
Welcome to the Q4 edition of Synapse Magazine, the official publication of the South African AI Association. This is the 25th publication of our ground breaking free to read magazine as we enter our 7th year of publication post launch way back in 2018. Time has flown since then and the AI landscape has radically change over that period. We wrapped the 7th Edition of AI Expo Africa held in Johannesburg on 31st October, sees largest gathering of Artificial Intelligence buyers and practitioners as AI Boom accelerates in Africa surpassing 2300+ delegates to cement our reputation as Africa’s Largest AI event. As with previous events, the core focus is real world AI applications that can be deployed by Enterprise, Government and NGO buyers with a diverse range of suppliers ranging from lead AI Hyper-scalers like Google to bleeding edge AI start-ups, educators, ecosystem builders and investors. The new CPD certified AI Skills workshop sessions were a massive success, and were run over 8 sessions to 1500 delegates and over 800 CPD certificates were issued by the SA AI Association. As with previous years, multiple VCs were attending and identifying candidate investments from the new AI start-ups exhibiting at the event. The current buzz around generative AI, data centres, national strategies, security and AI4Good were just some of the many themes our 60+ speakers and delegates were discussing. We were extremely happy to welcome the likes of Dr Simon See from Nvidia as a keynote speaker this year and a visiting delegation from the UN ITU AI4Good / UN DP community who ran a new AI4Good Pitching session at the show, won by SA Startup AI Diagnostics. A very big thank you to all the AI Expo Africa 2024 sponsors, exhibitors and partners that joined us.
DON’T FORGET Join the South African AI Association, to secure your FREE membership here >> https://saaiassociation.co.za/ and check out the SA AI Association’s “Start your Journey in Responsible AI” FREE training and resources >> https://saaiassociation.co.za/start-yourjourney-in-ai/ with content, training, guides, communities to help kickstart your adoption of Responsible AI.
We wish you all a happy Xmas break and see you all again in the new year for the 8th Edition of AI Expo Africa 2025 which will be joined by the inaugural UN ITU AI4Good Impact Africa event - watch this space for news on this in Q1 next year.
Happy Holidays from the whole team!!
PUBLISHER
AI Media Group
Web www.aimediagroup.co.za
EDITOR
Nick Bradshaw
GUEST EDITORS
Natasha Ochiel (Kenya)
Darlington Akogo (Ghana)
Naomi Molefe (South Africa)
Deon Van Zyl (Global)
EDITORIAL & ADVERTISING ENQUIRIES enquiries@aiexpoafrica.com
ACCOUNTS PAYABLE & ADMIN
Mia Muylaert mia.muylaert@aiexpoafrica.com
LAYOUT, DESIGN & PRINT
Karin Liebenberg
iCandy Design
Email: karinl@icandydesign.co.za
12 The true power of AI lies in purpose-built, enterprise AI solutions
14 Building AI Workforce for Tomorrow
17 Transforming Enterprise Document Generation with DocFusion
22 Google Search Trends
26 Bolstering AI for Good Innovation Across Africa
30 Mapping the Mind with AI: A Journey into the Challenges Unstructured Data
34 Unlocking the Future: AI and Cloud Solutions for South Africa’s SMBs
18 How we tackled modern-day bias with AI
50 Google Translate adds 25+ African languages
44 Gen AI: Easing the Burden of Consumer Decision-Making
2 AI Expo Africa – A Massive Success: Johannesburg Welcomed Largest Gathering Of Artificial Intelligence Experts
9 Unlocking employment opportunities for SA’s youth in the age of AI
24 Revolutionizing Your Desktop: AI Now Runs Locally
32 Preparing Data Centers for an AI-Driven World
36 Create Extra Contact Centre Capacity With AI-Powered Bots
38 Can AI Defeat Invoice Fraud in South Africa?
40 African Union: AU AI Strategy
42 Agentic AI: The next frontier in business automation
45 A new generation of African talent brings cutting-edge AI to scientific challenges
47 Lightweight Champ: NVIDIA Releases Small Language Model With State-of-the-Art Accuracy
51 AI challenge seeks ‘visionary ideas’ to tackle climate crisis
52 GMSA - AI Africa Report
53 The Forgotten: Leveraging AI to Promote Inclusivity and Equity for the Differently Abled
54 Bridging the Language Gap: Why African Language Models Matter
55 From RAG to Riches
56 United Siyafunda
58 Collaboration is Critical for Successful AI Development in Africa
60 South Africa shaping Responsible AI in the Military
Siyafunda
Africa’s largest Artificial Intelligence & Intelligent Automation event, AI Expo Africa 2024, was run at the Sandton Convention Centre, Johannesburg, South Africa 30-31 October 2024.
Now in its 7th year, the annual AI Expo Africa event united regional and international thought leaders, suppliers and start-ups to discuss and showcase the latest AI & Smart Tech innovations, from the biggest global brands to the smallest and newest local SMMEs.
The core focus of the show is real world applications you can deploy today with an audience comprised of Enterprise and Government buyers, deployment & service providers, start-ups, educators, NGOs and ecosystem builders.
Nonnie Kubeka, Executive Director of the Gauteng Convention & Event Bureau stated, “We were happy to welcome again the growing AI tech ecosystem to the Gauteng Region. As per last year AI Expo Africa 2024 hosted 48 SMMEs at the event. The gathering helped connect buyers and suppliers around one of the fastest growing sectors of the economy. AI is impacting all industries and certain aspects of society so its key we as a region, and a nation, have a focus on this sector as it’s a key pillar of our ICT activities and we want to showcase the local ecosystem to a regional and international audience.”
With a community of 50k+ practitioners, the show format is built upon the phenomenal success of the previous seven annual events that have cemented AI Expo Africa as the largest AI business gathering of its kind, attended by Enterprise buyers, Governments, practitioners, decision makers, international trade missions, vendors and SMMEs.
Dr Nick Bradshaw, Founder of AI Expo Africa & Chairperson of the South African AI Association stated, “This was our 7th Anniversary Edition and biggest show to-date, with over 1000 companies in the expo hall, many of which are local and regional SMMEs. The current buzz around generative AI, data centres, national strategies, security and AI4Good were just some of the many themes our 60+ speakers and delegates were discussing. We were extremely happy to welcome the likes of Dr Simon See from Nvidia as a keynote speaker this year and a visiting delegation from the UN ITU AI4Good community who ran a pitching session at the show."
The show was attended by some 2300+ delegates this year, and included new features like our CPD certified AI Skills workshop sessions and the SA AI Association Pavilion that showcased 10x local AI start-ups from South Africa. As with previous years, multiple VCs will be attending and identifying candidate investments from the new AI start-ups exhibiting at the event.
We are already working on the 8th Edition of AI Expo Africa, that will be held in Johannesburg 29-31 October 2025. Learn more: www.aiexpoafrica.com.
Press & Media: enquiries@aimediagroup.co.za
As AI takes over routine tasks in the workplace, the future job market demands innovation, adaptability and continuous learning – providing possibilities for the next generation to redefine work and thrive in a rapidly evolving landscape, writes Premlin Pillay, CEO of Tesserai and Group Executive of Mettus, a group of technology, data and analytics companies, with Tesserai leading the group’s data, analytics and artificial intelligence initiatives.
With the artificial intelligence (AI) market expected to see an annual growth rate of as much as 37.3% by 20301, its rise is transforming our world in unprecedented ways, revolutionising all aspects of life, and the workplace is no exception.
So much so that studies suggest that 64% of businesses consider AI a valuable tool for increasing overall productivity. This demonstrates that despite concerns about its impact on potential job displacement, it is here to stay.
With 77% of people expressing their concerns about AI impacting the job market, the World Economic Forum is more optimistic, stating that AI may create as many as 97 million new jobs.
To remain relevant amidst this rapid technological shift, South Africans,
particularly the youth about to enter the job market, should seek to adapt to the changing landscape, upskill themselves and adopt new skills to bolster their digital literacy, such as data analysis and machine learning.
AI skills are increasingly in demand. Young people proficient in AI technologies will have a competitive edge in the job market, making them more attractive to employers. Moreover, an understanding of AI can empower young people to innovate and create new business opportunities which can lead to the development of AI-driven startups and solutions, and contribute to economic growth.
Despite common perceptions, the advent of AI is not limited to the tech industry and is applicable across multiple industries, including healthcare, finance, education and agriculture, amongst others. Embracing AI opens up diverse career paths across various sectors.
The pace of technological change means that continuous learning and adaptability are crucial. Embracing AI fosters a mindset of lifelong learning, which is essential for career longevity and success.
As a career opportunity promising a wealth of possibilities for those with the right skills, mindset and expertise, there are diverse pathways South African youth can explore to maximise their potential for success.
One prominent area of career growth lies in data science. With the proliferation of data across industries, there is an increasing need for skilled data scientists who can extract meaningful insights from vast datasets. These professionals play a crucial role in informing decision-making processes and driving business strategies through datadriven analysis.
Similarly, machine learning engineering is another thriving field within AI. Machine learning engineers are responsible for designing and deploying AI models that can make predictions or decisions based on data.
Their expertise is essential in developing algorithms, training models, and optimising performance to achieve desired outcomes.
In addition, specialised AI fields like natural language processing (NLP) and computer vision offer diverse career opportunities.
As AI continues to penetrate every aspect of the workplace, it brings with it an increasing need for educators and trainers to impart AI skills and knowledge to the next generation of professionals, ensuring a skilled workforce ready to navigate the challenges and opportunities of the AI-driven future.
To this end, Tesserai has hosted an internship programme over the last three years to address the rapid growth of the AI industry, which often outpaces individual skills development. The programme takes cohorts of graduates from a range of disciplines and puts them through an intensive 12-month programme coupled with actual work experience.
Successful graduates from the programme are then placed across the Mettus group. The initiative aims to build capability in the data, analytics and AI domains, while ensuring a sustained advantage in the AI landscape through the cultivation of skilled professionals.
As South Africa’s youth stand on the precipice of the AI revolution, they hold the power to shape their destinies and redefine the future of work. By embracing innovation, fostering adaptability, and committing to continuous learning, they can unlock the boundless opportunities that AI presents and forge a future of prosperity and success.
Reference: Website [Internet].
Available from: https://www.forbes.com/ advisor/business/ ai-statistics/
/ By Len Pienaar – Managing Director at AdvanceGuidance /
In today’s rapidly evolving business landscape, integrating artificial intelligence (AI) into core business operations is essential for long-term success. Unlike generic AI services, enterprise AI solutions should be tailored to efficiently automate complex, repetitive tasks with predictable output quality. These systems are designed to continuously learn from data, refine their own operations, and enhance performance with a high degree of autonomy, often without direct human intervention.
The enterprise AI landscape is experiencing a rapid evolution, marked by significant growth in adoption, investment, and integration across various industries. The global AI market is projected to reach $407 billion by 2027, a substantial increase from past years, indicating the increasing integration of AI technologies across diverse sectors such as finance, telecommunications, healthcare and manufacturing.
Indeed, AI’s ability to analyse vast amounts of data, provide actionable insights, and perform complex tasks is revolutionising enterprise operations. This expansion
reflects not only the versatility of AI tools but also their growing necessity in maintaining competitive relevance in a data-driven world.
According to research commissioned by IBM, 42% of enterprise-scale organisations use AI actively in their businesses. With early adopters leading the way, 59% of responding enterprises are already working with AI, with plans to accelerate and increase their investment in the technology. The proof is real with early adopters, who have overcome barriers in deploying AI, continuing to make further investments, proving they’re already reaping benefits from using AI.
“The true power of AI for business lies not in generic AI solutions but in a strategy that deploys purpose-built enterprise AI solutions tailored to the unique needs and challenges of individual businesses,” explains Len Pienaar, Managing Director at AdvanceGuidance.
In his journey as a CEO across sectors as diverse as banking, telecommunications, and media, he’s witnessed firsthand the transformative potential of data and enterprise AI solutions.
Enterprise AI is expected to grow at an impressive 37.3% CAGR from 2024 to 2030. With the growth driven by the advancements in machine learning and natural language processing (NLP), AI is predicted to become more accessible, effective for businesses and critical in enhancing existing business models.
The importance of AI in shaping future business strategies is made apparent by the fact that 83% of companies place AI at the forefront of their strategic planning. This indicates the shift in business mentality, recognising the potential of AI solutions to drive innovation, efficiency, and profitability. It’s clear that AI is no longer an optional tool but a fundamental component in corporate strategy and decision-making processes.
“At AdvanceGuidance, it’s about more than just data. Once we’ve understood the business problem and analysed how to solve it, the real alchemy of what we do lies in how we use AI to provide unprecedented insights
that are implemented to rapidly generate business results for our clients.” shares Pienaar.
“We specialise in helping large corporations design, implement, and operate enterprise AI and machine learning solutions with deep data science and management expertise,” adds Pienaar.
“I firmly believe that AI will impact businesses and society more than digital computers and mobile devices combined. In the short-term, AI may be overhyped, but few truly grasp the long-term impact that AI will have on businesses”.
Says Pienaar, “The expansion of AI at the enterprise level is due to the increasing accessibility of AI tools, the significant drive in the automation of key complex processes, coupled with increasing amounts of AI being embedded into commercially available business applications.”
He cautions, however, that the road to successful enterprise AI implementation is fraught with challenges, particularly in highly regulated industries where compliance, data privacy, and security are paramount.
The challenges of implementing AI in complex environments are significant, including those around data privacy, scalability and managing stakeholder expectations. And that’s without considering hiring employees with the right skill sets, data complexity, and various ethical concerns that continue to prevent businesses from adopting AI technologies into their operations.
Drawing on his experience in executive positions at FNB, MTN, Media24, Clickatell and CellC, don’t miss Len Pienaar as he presents at The AI Expo Africa, the largest Enterprise AI, RPA & Smart Tech trade event in Africa, from 30th-31st October 2024, exploring ‘Why a tailored enterprise AI strategy is critical for businesses wanting to thrive in the AI age.’
Using a case study of a proof-of-concept project for a global consulting firm, he’ll demonstrate how purpose-built enterprise AI solutions can enhance operations, ensure
compliance, and provide a competitive edge to businesses of the future.
He’ll also address the challenges of implementing AI in complex environments, including data privacy, scalability, and managing stakeholder expectations.
The future of AI in business lies in its ability to enhance, improve, or even replace complex, repetitive tasks traditionally performed by skilled professionals. This shift will fundamentally change operations’ cost structure and efficiency across industries, setting the stage for long-term success.
It’s a fact that any organisation that regularly engages in proposal writing, RFP responses, or other repetitive, high-value tasks stands to already benefit from AI. By shifting these tasks to purpose-built AI solutions, businesses can free up their most valuable resources—human expertise—for more strategic activities.
Companies that embrace this challenge of transformation will be able to create growth for their business and value for their customers, while building a workforce that’s adaptable and agile for the future.
AI at work is no longer a question, it’s arrived. By embracing a skills-first, human-centric, learning-led approach, and strategically implementing AI across functions, companies can begin to move from AI buzz to AI breakthroughs.
As we move forward, it’s imperative to consider how AI can be integrated into every aspect of business operations, from data management to day-to-day activities and even decision-making. Today’s decisions about your AI strategy will shape your competitive landscape tomorrow. Don’t just stay current; it’s up to you to prepare for the future by incorporating enterprise AIs into your corporate strategy.
Pienaar has enjoyed a lifelong interest in how things work and how they can be improved, with a background in engineering and business administration. With over 20 years’ experience in senior leadership positions, he is equally comfortable in large multinationals and start-ups.
Pushing boundaries, a determination to succeed, and assembling and leading exceptional teams are the building blocks of Pienaar’s career. They are underpinned by exceptional analytical ability. His focus on tech innovation and customer adoption, from conception and strategy to execution and delivery, has enabled him to launch and commercialise numerous industry-first innovations, for which he has received local and global recognition. Len Pienaar became the Managing Director of Advance Guidance in July 2024.
Designed
What if enterprise document processes didn’t have to be complicated, time-consuming, and errorprone? With the right technology, these challenges can be simplified, allowing enterprises to focus on what truly matters.
The document demands placed on enterprises – especially within the financial services sector – are immense. Documents are not just static records; they are the lifeblood of business operations, capturing critical transactions and information across the enterprise.
However, the sheer volume and complexity of these documents often present formidable challenges to operational efficiency, security, and compliance. This is where DocFusion stands out as an enterprise-grade solution that meets these challenges head-on.
DocFusion is a comprehensive document generation and automation platform designed to handle the vast volumes and complexities of enterprise documents. Capable of processing thousands of documents per second, DocFusion ensures that businesses can scale seamlessly, meeting the demands of any operation.
This level of performance enables speed and efficiency, while also ensuring accuracy, compliance, and security at every step of the document lifecycle.
Tailored for the Financial Services
While DocFusion is versatile enough to serve various industries, it has seen its most widespread application and success in the financial services sector.
Financial institutions deal with some of the most sensitive and regulated information in the world. Compliance is non-negotiable,
and any deviation can result in severe financial and reputational risks.
DocFusion’s platform is designed to integrate effortlessly with existing enterprise systems, pulling in the freshest data directly from the source. This integration ensures that every document generated is accurate, compliant, and ready for immediate use.
The challenge in industries like banking and insurance lies not only in document volumes, but also in document complexity, and the disarray that often accompanies it. For example, conside r the process of generating a comprehensive insurance policy document or a detailed loan agreement. These documents require input from multiple departments—such as underwriting, legal, compliance, and finance—and must adhere to strict regulatory standards and companyspecific guidelines.
The complexity arises when data from these various departments needs to be accurately compiled into a single, compliant document. Any errors or inconsistencies could lead to significant legal or financial repercussions.
DocFusion addresses this complexity by dynamically sourcing and consolidating business data from disparate systems, using approved templates that are designed to meet the stringent requirements of industries like banking and insurance. This ensures that every document, whether it’s an insurance policy or a loan agreement, is generated with the highest level of accuracy, compliance, and efficiency.
DocFusion’s strengths are underscored by its capabilities in five key functional areas:
1. Document generation: Dynamic creation of documents in various formats using intelligent templates.
2. Electronic signing: Securely capture digital signatures, ensuring compliance and legality.
3. Version control: Maintain a secure audit trail for all document templates, complete with approvals and access controls.
4. Document workflows: Ensure documents are available when needed, maintaining content integrity from creation to output.
5. API integration: Enable seamless connectivity with enterprise systems, allowing DocFusion to integrate efficiently with existing workflows, often through low-code or no-code interfaces.
What sets DocFusion apart is its ability to turn complexity into simplicity. While other document automation products may add layers of data distribution that complicate business processes, DocFusion integrates
...continues on page 20
/ Ekow Duker – Co-founder, The AI Shop /
In an age where technology is reshaping every aspect of our lives, The AI Shop and Africa No Filter have introduced a powerful new tool designed to combat bias in writing - The Africa Bias Buster. This AI-powered solution is not just another technological innovation; it’s a direct response to the persistent stereotypes that have long distorted the portrayal of Africa in the media and beyond. The Bias Buster helps writers identify and eliminate unconscious bias in their work, creating opportunities for more balanced, accurate storytelling about Africa.
...continues on page 20
MANAGE Unorganised data
DATA
- Invoice Processing
- Contract Processing
- Credit Applications
- Know your Customer (KYC)
- Policy Schedule Processing
- Communication Classification
- Claims Processing
- CV Processing - Automated Claims Processes - Survey Verbatim Classification - Document Classification
- Handwriting Recognition
- Photo and Signature Recognition
- Sentiment Detection
- Knowlege Management
RAG - Payslip, Bank Statement
...continued from page 18
Since the launch of ChatGPT in November 2022, much has been written about the risks of algorithmic bias—ranging from denying people job opportunities to wrongfully convicting the innocent. However, there’s another type of bias that’s been around much longer, and it’s one Africans know all too well.
In the 18th century, supporters of the transatlantic slave trade justified their position by characterizing Africans as inferior beings. Edward Long, one of the era’s most vocal advocates, described Africans as “proud, lazy, deceitful, thievish, addicted to all kinds of lust”. Despite never setting foot in Africa, he marvelled that, even after long periods of colonization, Africans demonstrated no “appreciation for the arts or any inventive ability”.
This type of bias has persisted across generations. Even two centuries later, Che Guevara echoed Long’s sentiment, describing the African as “indolent and lazy”, and who “spends his money on frivolities, whereas the European is forward-looking, organized and intelligent.”
Thankfully, such virulent stereotypes are not as overt as they once were. But while political correctness may have turned down the vitriol, such stereotypes haven’t really gone away. We still encounter well-meaning writing that, in subtle ways, implies Africans lack agency and are dependent on Western saviours.
Let’s look at an example: In a blog post subtitled ‘Humanitarian heroes take up the call to improve the welfare of others’, a well-known development agency writing about its work in South Africa, depicts humanitarian organisations and people who donate to charities as ‘heroes’. While the text seems harmless at first glance, a closer look reveals that it downplays the role of local heroes and minimizes the agency of African communities in driving their own progress.
It’s exactly this sort of bias that Africa No Filter, a donor collaborative committed to changing outdated perceptions of Africa, seeks to weed out. Their mission is to promote stories that go beyond stereotypes and offer a more nuanced and accurate portrayal of the continent. As part of this mission, Africa No Filter envisioned a toolpowered by AI - that would help storytellers create more balanced narratives about Africa. Abimbola Ogundairo, Advocacy and Campaigns Lead at Africa No Filter imagined it like this: “An AI-powered learning tool that analyses your writing to identify stereotypes and provide actionable feedback to help you tell more impactful stories, all in real-time!”. They called this tool The Africa Bias Buster and asked The AI Shop to build it with them.
“At The AI Shop, our job is to make enterprise workflows run better,” said the coFounders of The AI Shop, Ekow Duker and Simon Stewart. “We saw this as a critical project because bias, especially when aimed
“ Thankfully, such virulent stereotypes are not as overt as they once were. But while political correctness may have turned down the vitriol, such stereotypes haven’t really gone away. We still encounter wellmeaning writing that, in subtle ways, implies Africans lack agency and are dependent on Western saviours.”
at entire populations, can have insidious, long-term effects. It was an honour to co-create this tool with Africa No Filter and demonstrate how AI can be a force for good.”
You can try out the Africa Bias Buster yourself at https://africanofilter.org/home Just type in or upload a piece of writing and sit back while the bias baster analyses the text, assigns a score out of 5, and gives you suggestions on how to get rid of any subtle or overt biases that may be present in the writing.
This novel collaboration between Africa No Filter and The AI Shop successfully developed and refined an AI-powered tool designed to identify and address biases in narratives about Africa. By leveraging advanced language models and incorporating user feedback, we created a robust solution that empowers writers, journalists, and other stakeholders to produce more balanced and truthful portrayals of the continent. This initiative not only supports Africa No Filter’s mission to promote better storytelling about Africa but also demonstrates how cutting-edge technology can be used to address a centuries-old problem. Edward Long would not be pleased.
“ In the 18th century, supporters of the transatlantic slave trade justified their position by characterizing Africans as inferior beings. Edward Long, one of the era’s most vocal advocates, described Africans as “proud, lazy, deceitful, thievish, addicted to all kinds of lust”. Despite never setting foot in Africa, he marvelled that, even after long periods of colonization, Africans demonstrated no “appreciation for the arts or any inventive ability.”
...continued from page 17
seamlessly with existing systems, ensuring that the freshest, most accurate data is always used. This approach improves efficiency and eliminates costly errors and the need for extensive system integration projects.
What’s more, DocFusion’s template management system centralises document templates, ensuring that all users access the most current versions, with robust version control, access permissions, and seamless integration capabilities.
DocFusion is not just a solution for today; it is built to evolve with the needs of large enterprises. We recognise the transformative role of AI in shaping the future of document generation. Our vision for AI is to enhance human capabilities, creating a synergy between artificial and human intelligence that reduces tedious tasks and boosts productivity, all while maintaining the essential element of human interaction.
As DocFusion sponsors Africa’s biggest B2B AI Expo this October, we invite you to explore how our enterprise-grade document generation and automation solutions can transform your business operations.
Visit our booth to see DocFusion in action and discover how we can help your organisation overcome the challenges of document complexity, ensuring operational efficiency and compliance at any scale.
DocFusion Website: www.docfusioncloud. com
DocFusion Blog: DocFusion: True enterprise-grade document automation
DocFusion Blog: The magic of low-code document generation
DocFusion Blog: Template Management: The key to successful document generation
ExampleRDHx @ 90kW per rack
✔ Our Powercube AI orientated HPC Containerised Modular Data Centre solutions deliver flexible, scalable,Tier 2, 3 or 4 - Hybrid or pure HPC compute environments.
✔ We can integrate DX. RDHx. DLC. and/or IMMERSION technologies to suit your specific requirements/choice of vendor.
✔ Design, build, plug-in and play with zero disruption to your existing data centre.
✔ Integrate your preferred technology partners.
Custom Powercube Containerised Modular Data Centres are fast becoming the choice for mainstream RDHx, DLC and Immersion technologies.
From 1 to 38 racks in a single Powercube container footprint. Co-join multiple Powercubes to create larger flexible DC environments.
Artificial Intelligence (AI) was searched more than ever in 2024 globally and in South Africa, with South African users searching for the term 120% more than this time last year.
Search trends in South Africa related to AI continuesearches for “how to use AI” nearly doubled (+70%), while searches for “what is AI” have increased by 20%.
With interest in AI growing, South Africans are looking to build AI skills and qualifications - with searches for “AI course” and “AI jobs” nearly doubling.
22 August 2024: New search trends released by Google today show that people in South Africa are searching for AI more than ever before.
Search interest in “AI” has reached a record high around the world and in South Africa, where searches have
more than doubled (+120%) since the same period last year. Searches in South Africa for “what is AI” also increased by 20%, while searches for “how to use AI” increased by 70%.
South Africans are looking to use AI to build their careers and creativity - with trending searches including “AI image generator”, “AI writer”, “online AI chat” and “logo maker AI” - all of which were ‘breakout’ searches in 2024, growing over 5,000% in interest since this time last year.
The research also shows that South Africans are keen to grow their AI skills and qualifications. Searches for “AI jobs” and “AI course” both nearly doubled (+80% each). Searches for “AI and business” doubled (+100%) while searches for “AI and startups” increased by 30%.
Aligned with its commitment to AI as a force for positive societal change, Google offers free AI training programs to Africans through its Hustle Academy. This initiative, designed for entrepreneurs, includes modules specifically tailored to teach participants how to leverage AI to improve their businesses. Since its launch in 2022, Hustle Academy has empowered over 10,000 businesses. Additionally, Google supports promising AI-focused startups in Africa through its Google for Startups Accelerator Africa program. This program, which assists startups tackling African challenges, selected 10 AI-driven ventures for its 2024 cohort.
Search interest in cybersecurity and AI also doubled across South Africa (+100%), while people showed a clear interest in the future of AI for science. Searches for AI and doctors doubled (+110%), while searches for AI and medicine increased by 50%, searches for AI and hospitals increased by 60% and searches for AI and healthcare increased by 20%.
Paul Mayanja, Acting Google SA Country Director,: “AI has the potential to create opportunities - from the extraordinary to the
everyday - for everyone: bringing new waves of innovation, social and economic progress. So it’s no surprise that the people of South Africa are already looking to make the most of this transformational technology - finding ways to learn more AI skills to boost their career, creativity, and society.”
The results come months after Google’s global survey, “Our Life with AI”, which showed that 88% of South Africans were interested in learning more about AI. The survey also showed that 60% of South Africans were excited about AI, with 76% agreeing that “society as a whole” is likely to benefit from AI.
Notes to editor:
All trends, unless stated otherwise, are measured from January 1 - July 15th 2024. All percentage increases are in comparison to January 1 - July 15th 2023.
About Google’s work with AI: Google began working on AI in 2012, first by training computers to recognise cats on YouTube. We released TensorFlow in 2015, our internal machine learning and AI software library. In 2016 we published our AI Principles for the first time: our public outlining of how we pursue AI boldly, responsibly and collaborativelyas well as AlphaGo and Transformer. Earlier this month, at Made by Google, we shared how we’re bringing the best of Google AI to more people and more phones with updates across Gemini, Android and the Pixel portfolio. See the announcements here Google is also using AI to address some of society’s biggest challenges. In Ghana, Google’s AI Center is working in collaboration with InstaDeep and the UN to better detect locust outbreaks and implement control measures. While in Kenya, Google is working with Jacaranda Health to improve maternal care.
Pinnacle has announced a significant expansion of its artificial intelligence (AI) portfolio, introducing a range of new services and consulting capabilities designed to help businesses navigate the rapidly evolving AI landscape.
This strategic move underscores Pinnacle’s commitment to providing comprehensive AI solutions that not only meet the demands of today’s market but also anticipate the needs of tomorrow.
With AI becoming an increasingly critical component of business strategy, Pinnacle’s expanded offering aims to support organisations in leveraging AI to drive innovation, enhance efficiency, and improve decision-making processes. The new services are tailored to assist businesses at every stage of their AI journey, from initial planning and strategy development to full-scale implementation and ongoing optimisation.
A key component of Pinnacle’s expanded AI offering is its consulting services, which provide businesses with access to a team of AI experts who bring deep industry knowledge and technical expertise. These consultants work closely with clients to understand their unique challenges and objectives, delivering customised AI strategies that align with their business goals. This personalised approach ensures that organisations can maximise the benefits of AI while minimising potential risks and disruptions.
In addition to consulting, Pinnacle is offering a suite of managed AI services, including data management, machine learning model development, and AI-powered automation solutions. These services are designed to streamline the adoption
of AI technologies, allowing businesses to focus on their core operations while Pinnacle handles the complexities of AI integration and management.
Pinnacle’s CEO, Tim Humphreys-Davies, highlighted the importance of this expansion, stating, “As AI continues to reshape industries, it’s crucial for businesses to have the right tools and guidance to harness its full potential. Our expanded AI services and consulting capabilities are designed to provide that support, helping our clients achieve their strategic objectives through intelligent, data-driven solutions.”
The expansion also reflects Pinnacle’s ongoing investment in AI research and development, ensuring that its offerings remain at the forefront of technological innovation. By combining cutting-edge AI technology with expert consulting and support, Pinnacle is well-positioned to help businesses of all sizes and sectors successfully integrate AI into their operations.
As AI adoption continues to accelerate across industries, Pinnacle’s enhanced services and consulting expertise will be instrumental in helping businesses not only keep pace with the competition but also lead the way in their respective markets. This latest development solidifies Pinnacle’s role as a trusted partner for organisations seeking to navigate the complexities of AI and unlock new opportunities for growth and success.
For more information, contact us: www.pinnacle.co.za | +27 (0) 11 265 3000
/ By Deon Van Zyl /
Artificial intelligence (AI) has become the new focal point of innovation in desktop computing. Traditionally AI tools were tied to cloud platforms or standalone applications. Currently AI is being woven directly into operating systems. Fundamentally altering how users interact with their devices. From smart assistants to automated workflows, these integrations aim to streamline productivity while enhancing accessibility and offering powerful security features. However, they also raise important questions about security, privacy and resource allocation.
The integration of AI into desktop computing offers numerous benefits that promise to transform how we use technology. By automating repetitive tasks, such as drafting emails or organizing files, users can focus their energy on other assignments. This is especially evident with systems like Copilot, which integrates across multiple applications to streamline workflows. Similarly, AI's ability to personalize user experiences is another major strength.
AI tools like Google Jarvis can provide contextually relevant recommendations. Additionally, the integration of AI within operating systems could reduce the reliance on external software which in turn could potentially save costs for the user. Nevertheless, these advantages are tempered by several drawbacks. The most immediate issue is the resource demand of AI-driven features, which can significantly influence older devices. Another concern is the potential over-dependence on AI for everyday tasks which also requires time and effort to learn. This could lead to a decline in users fundamental computer skills.
Taking into consideration these general pros and cons of AI integration it is beneficial to investigate the new AI features coming to desktop computers. For instances, Generative assistants which is an AI assistant that provides initiative support across various tasks such as drafting documents to automating desktop workflows. As well as Automation tools that can predict user behavior and recommend or automate actions, streamlining repetitive tasks and improving efficiency. In the “
The integration of AI into desktop computing offers numerous benefits that promise to transform how we use technology. By automating repetitive tasks, such as drafting emails or organizing files, users can focus their energy on other assignments. This is especially evident with systems like Copilot, which integrates across multiple applications to streamline workflows. Similarly, AI's ability to personalize user experiences is another major strength.
AI tools like Google Jarvis can provide contextually relevant recommendations. ”
“
AI integration offers better security and privacy as it is in the operating systems. AI can more actively monitor device vulnerabilities, identify suspicious behavior, and apply protective measures before issues escalate.
On-device data processing minimizes the risks associated with transmitting sensitive information to the cloud. This local processing approach offers a higher level of privacy by ensuring user data stays within their devices. However, the reliance on AI systems for security also brings challenges because many AI features require significant amounts of user data to function effectively. ”
creative space, improved capabilities such as powerful photo and video editing tools. To be more inclusive, Accessibility enhancements has for example real-time transcription, screen reading, and gesture recognition.
Features announced by major players
Apple introduced Apple Intelligence which is a suite of generative AI features that is now available on iPhone, iPad, and Mac. With deep integration into apps like Messages, Mail, and Photos. Apple intelligences enhance user experiences with tools like auto-summarization and photo editing. This includes generating creative text responses and automatically organizing media. Microsoft, another major player, announced updates for Windows 11 and Copilot+ PCs on October 1, 2024. Introducing a range of AIdriven features to enhance productivity and creativity. Among the key additions is Recall, which allows users to locate past files and interactions securely with user-controlled privacy settings. Click to Do is aimed at automating task and streamlining workflows. Anthropic unveiled Claude's computeruse capabilities which allows AI to interact directly with the machine and software through virtual keyboard and mouse actions. This feature is currently in public beta that supports tasks like navigating screens, executing commands, and responding to user prompts. The last well-know player is Google. Google Jarvis transforms the desktop experience by automating environments and managing everything
from file organization to real-time task coordination. Initially aimed at ChromeOS, this may soon be able to install on other desktop operating systems as well.
AI integration offers better security and privacy as it is in the operating systems. AI can more actively monitor device vulnerabilities, identify suspicious behavior, and apply protective measures before issues escalate. On-device data processing minimizes the risks associated with transmitting sensitive information to the cloud. This local processing approach offers a higher level of privacy by ensuring user data stays within their devices. However, the reliance on AI systems for security also brings challenges because many AI features require significant amounts of user data to function effectively. Therefor raising concerns about how this information is stored and whether it might be exploited. While challenges do remain, the potential of AI-powered desktops is undeniable.
Deon is a sophisticated technical IT professional with a solid history of effectively bridging the gap between Development, Security, Automation, and Artificial Intelligence. His track record of over 25 years, has a footprint that spans major corporations, academic institutions, and government.
Artificial Intelligence (AI) is rapidly transforming industries. Innovators across Africa are developing homegrown solutions to challenges in healthcare, agriculture, education, and governance. From Nairobi’s burgeoning tech hubs to Lagos’ vibrant startup scene, African nations are not only adopting AI but also shaping it to suit local needs. Yet, this rapid growth requires continued investment to ensure equity and access to the infrastructure and capacity that support continued AI innovation.
At AI for Good, we are committed to nurturing innovation across the continent. Through challenges and support for AI startups, granting cloud computing credits, supporting scholars, and working with local partners, there are numerous success stories coming out of Africa when it comes to AI for social good projects.
As the leading United Nations (UN) platform for AI that supports the Sustainable Development Goals (SDGs), we are organized by the International Telecommunication Union (ITU) and supported by 47 UN Agencies.
Our AI for Good Innovation Factory is the UN-led pitching platform to help startups grow and scale their innovative AI-powered and SDG-driven solutions by connecting them to potential business partners, investors, governments, big IT companies, and more. For those who are seeking groundbreaking solutions from startups, it is an excellent platform to explore opportunities to meet promising ventures and collaborators. Several startups have been able to scale their solutions after winning our prize.
In 2022, OKO Finance, an innovative crop insurance company, won our Innovation Factory prize for work in Mali and Uganda. Since then, it has continued to grow and is now operating in seven countries. OKO has an impressive double impact: On the one hand, they contribute to financial inclusion, bringing insurance, access, and credit to people previously deprived of these basic services. On the other hand, they contribute to climate resilience by providing for better prediction. Watch our session “Meet the Africa-based AI Start-up Innovations” on YouTube to learn more.
This year, we are hosting another Innovation Factory for Africa organized in partnership with AI Media and open to all African-based startups.
Co-hosted by AI Media, a regional media and advisory consultancy focused on the emerging AI opportunity in Africa, this global startup pitching session aims to identify the most promising African startups using AI to solve sustainable development challenges. If you are an African startup with novel AI solutions to shape a sustainable future, don’t hesitate to apply here by 20 September
In addition to supporting startups, AI for Good has launched a new program to recognize scholars and projects making an innovative impact on SDGs.
In the recent report, Innovate for Impact, students working on issues of sustainable development were encouraged to apply with their projects to win awards and recognition. Some of the chosen projects from Africa include: “Using AI to Reduce the 6G Standards Barrier for African Contributors.” This project from Nigeria’s Federal University of Technology, Minna, is helping to bridge the digital divide with AI solutions.
In Tanzania, the government is developing an AI-powered system for enhancing transparency and accountability in public procurement and project monitoring. The Prevention and Combating of Corruption Bureau developed Natural Language Processing (NLP) and Large Language Models (LLMs) to analyze procurement projects with open-source data, in an attempt to limit corrupt and unethical practices.
In Libya, Leapchat AI developed an AI chatbot to support families of missing persons. It was a collaboration between an
AI chatbot platform operating a corporate social responsibility (CSR) program and Libyan non-governmental organizations (NGOs) to provide support to individuals affected by the Derna flooding crisis.
The report contains over fifty innovative use cases where AI is being used to improve lives.
Lastly, there are some great outcomes when AI for Good collaborates with local partners. For several years, AI for Good has been working with Zindi to find the best data-driven solutions through data science competitions.
Zindi hosts the largest community of African data scientists, working to solve the world’s most pressing challenges using machine learning and AI. The International Telecommunication Union (ITU) partnered with Zindi to develop solutions to important global challenges through a series of data science competitions hosted by AI for Good. In these challenges, hundreds of data scientists from around the world are building complex machine learning models to help map potential climate disasters, predict and reduce energy consumption, and address the Sustainable Development Goals (SDGs) in innovative ways.
Here are some of the ways that these projects are making a difference:
Disaster prevention and supporting agriculture around the world
Landslide susceptibility mapping can help local authorities plan and implement sustainable development measures, reduce the risk of landslides, and ensure the safety of communities living in high-risk areas. In this challenge, creating a hazard map with AI allows governments to keep their citizens safe from disasters. See more here.
Accurate and up-to-date crop maps are essential for agriculture as well as other relevant fields, such as natural resources, environment, health, and sustainability.
Cropland extent maps are the basic products that allow for practical agricultural applications. Through this competition, Zindi and ITU are enabling a more precise and comprehensive understanding of agricultural landscapes worldwide.
Zindi and ITU have created a data science competition to use machine learning to produce air pollution susceptibility maps in Milan, Italy, which will support local government decision-making to improve the public health and resilience of the city. Zindi believes that through studying our available data and creating predictive models, the world can become a better and safer place for all.
In a world where energy use is responsible for three-quarters of fossil fuel consumption, there is a desperate need for ways to reduce this. Using AI, Zindi’s community of data scientists are finding new ways to reduce energy consumption in 5G networks.
Managing faults in complex telecom networks is an enormous and demanding task. Being able to predict faults in Radio Access Networks (RANs) will lead to reduced costs and improvements in network uptime. ITU and Zindi believe that a platform like Zindi benefits both the organizations providing the data and the participants of the challenge, who develop real skills by working on real problems.
Through these ITU AI/ML Challenges, select winners from the 2023 problem statements were invited to participate in the 2024 AI for Good Global Summit. These individuals had the opportunity to present their solutions during the summit’s workshops, sharing their insights and experiences with a global audience.
“The ITU AI/ML Challenges provide an invaluable platform for ML scientists to hone
their skills and address real-world problems at the cutting edge of technological advancement. My participation exposed me to a broad range of international challenges, fostering a deeper understanding and expertise on the critical applications of AI/ML in various fields, including 5G/ Networking. The series have made me more fit for both the industry and academia than ever before. Therefore, if anything, the ITU AI/ML challenges should keep expanding and collect more problem statements to expose the community further,” said Ndabuye Gideon, one of the participants.
Her testimonial underscores the significant role the AI for Good ITU AI/ML Challenges play in advancing participants’ careers and contributions to the field of AI and Machine Learning.
In addition to our work with Zindi, AI for Good is partnering with Seeed Studio for local programs at the Malawi University of Science and Technology (MUST) to support AI innovators. A “Summer School Workshop on TinyML” brought around 50 students to help develop capacity that could enable local solutions in low-cost settings throughout the month of August.
TinyML is a technology that allows machine learning models to run on lowcost, low-power microcontrollers. TinyML’s extremely low bandwidth and energy requirements make it uniquely suited for regions with limited access to reliable energy and computing infrastructure.
This technology plays a significant role in advancing the Sustainable Development Goals (SDGs) and enabling new applications in fields such as healthcare, agriculture, environmental monitoring, and conservation.
TinyML has attracted significant interest from researchers, developers, and industries for its potential to enable innovative applications in healthcare, agriculture, transportation, conservation, smart homes, and more. Though currently restricted in reach, TinyML intersects topics across computer science and engineering curricula, making it an impactful educational tool.
This hands-on workshop focused on TinyML applications relevant to Malawian students and researchers, providing training on commercially available hardware optimized for embedded ML deployment. By making TinyML more accessible, this workshop empowers researchers to develop localized solutions that benefit their communities.
It is estimated that digital technologies directly benefit 70 percent of SDG targets, and these collaborations are showing how advanced technologies like AI can come from anywhere in the world to solve these global challenges.
Through these ongoing collaborations, AI for Good is helping to support everyone who wants to be a part of the solution. To learn more about AI for Good and to get involved in ongoing projects in Africa, join us!
Data
Data
Reporting
Data
/ By Tyler Donaldson, Lead Data Scientist, OLSPS /
Recent developments in large scale LLMs (large language models) are supercharging behavioural analytics in new ways. OLSPS has been at the forefront of mapping in-situ workplace mindsets and behaviours for the past 15 years. Building rich, situationally aware behavioural frameworks that explain as much as 40% of bottomline employee performance. Now, using free-text data and 15 years of insights, OLSPS is pushing the envelope on a situationally aware behavioural AI that can do everything from coaching individuals to informing company strategy, structure, and culture.
The firepower in our arsenal is the rapidly evolving space of LLMs. These LLMs are semi-intelligent in that they already understand languages and language structure, and even, to an extent, the general basis of human mindsets and behaviours. It is possible to use these LLMs as a base model which can then be ‘fine-tuned’ by training it further on very specific data and prompting it with context and limitations. They are easy to use, affordable, and readily available. The real challenge in implementing these LLMs, which isn’t limited to this use case, is the data.
The data preparation for ‘fine tuning’ one of these LLMs is not a trivial task.
“ Data labeling and annotation is another labor-intensive aspect. This process can be slow and expensive, requiring human expertise to ensure accuracy and consistency. Mislabeling or inconsistent annotations can degrade the model's learning process. ”
Firstly, data quality and relevance are paramount. Poor data quality, or irrelevant data, can lead to suboptimal model performance, which could misrepresent customer interactions or make inaccurate predictions.
The volume and diversity of data also pose significant challenges. A model trained on too little or too narrow data might excel within those parameters but fail when encountering new scenarios or create discriminatory biases in the LLM.
Data labeling and annotation is another labor-intensive aspect. This process can be slow and expensive, requiring human expertise to ensure accuracy and consistency. Mislabeling or inconsistent annotations can degrade the model's learning process.
Cleaning and preprocessing of data involves normalizing text, handling missing data, and ensuring privacy by removing personal identifiers. This step is crucial to avoid introducing errors or biases into the model, which could have legal or ethical implications.
From a computational standpoint, there’s a balance to strike between model performance and resource use. Fine-tuning demands significant computing power, and businesses must manage this cost against the expected enhancement in model accuracy.
Ethical and legal considerations are vital, particularly around inherent bias in the data. Ensuring the model does not perpetuate or introduce discriminatory biases requires careful data curation.
Context of the data is also significant, to avoid the LLM misinterpreting the meaning of certain data. The often requires business discovery, and an understanding to the data sources, usages, and systems.
Lastly, multilingual and cultural adaptation along with continuous learning to handle data drift are ongoing concerns. As customer bases or operational environments change, models need regular updates to stay relevant, which involves revisiting the data preparation phase.
In essence, for a business, the journey from raw data to a finely tuned LLM involves navigating through a maze of technical, ethical, and operational challenges. This process is where experienced data scientists are still required to effectively utilise these revolutionary LLMs in specific contexts.
For PROPEL, the behavioural analytics tool that OLSPS has been developing with Enjol for the past 15 years, the data problem is lessened by the fact that our data collection was designed in-house by data scientists from the outset, meaning most of the above challenges can be simply solved with the right know-how.
Beyond the data, our AI is agent based, and heavily prompted. The flow of decisions and integrations needs to be designed and integrated. Ongoing training, improvement, and model lifecycles need to be considered. As well as several other issues.
At the end of the day, the LLMs available now are incredibly powerful tools. They cannot, however, be blindly applied. Applied AI requires the experience and wisdom of applied data science. Failing to prepare the data correctly, curtail the scope of the LLM, prompt it concisely, or effectively curate it, can often lead to rouge, unhelpful LLMs and even lawsuits.
“ From a computational standpoint, there’s a balance to strike between model performance and resource use. Fine-tuning demands significant computing power, and businesses must manage this cost against the expected enhancement in model accuracy. ”
“ At the end of the day, the LLMs available now are incredibly powerful tools. They cannot, however, be blindly applied. Applied AI requires the experience and wisdom of applied data science. Failing to prepare the data correctly, curtail the scope of the LLM, prompt it concisely, or effectively curate it, can often lead to rouge, unhelpful LLMs and even lawsuits. ”
As artificial intelligence revolutionizes the business world—allowing companies to analyze and extract insight from data at lightning speed—it’s also transforming the data center. No longer can companies simply add another rack of multipurpose servers. They must equip their centers in a whole new way.
Moreover, AI workloads are more computationally intensive than traditional ones, meaning that gaining immense computing power while increasing efficiency has become more critical. Those needs will only grow as AI use soars in coming years and workloads become more complex. According to Grand View Research, the global AI market size is expected to increase at a 37.3% compound annual growth rate between 2023 and 2030.
Given the multifaceted nature of running AI, companies increasingly must view their data centers as “AI factories,” says Alok Srivastava, senior solutions manager at Supermicro, a leading NVIDIA partner and provider of AI-capable, high-performance servers, management software, and storage systems. The term “AI factory” was coined by NVIDIA founder and CEO Jensen Huang. It refers to a cohesive system explicitly designed to run AI workloads—or even a single AI application, such as generating human-like responses with a chatbot — with optimum efficiency.
“We traditionally think of factories as where manufacturing and mass production happens,” Srivastava says. “In an AI factory, you’re manufacturing intelligence: Raw data goes in, intelligence comes out.”
These systems blend the hardware components needed to run AI applications with the necessary software components, such as machine-learning frameworks, dataprocessing libraries, and scalable storage solutions. The value of this approach is that you’re creating an optimized system that can provide maximum performance for the tasks at hand, whether training generative AI models, deploying real-time inference at the edge, or accelerating high-performance computing applications.
“You gain efficiency and performance when you create a system around one workload instead of trying to serve multiple workloads simultaneously,” Srivastava says.
As companies create an AI factory, they must prioritize scalability—given that data volumes are only expected to grow exponentially—and flexibility to adapt to changing AI demands, he adds. This futureproofing can reduce long-term costs while allowing companies to stay ahead of the rapidly evolving AI curve.
Supermicro creates built-to-order AI systems by taking a building-block approach. It allows customers to tailor systems for their exact AI workloads, optimizing for factors like GPU density, energy efficiency, and I/O throughput. Those customized systems include the specific hardware components and the software and applications needed to run that particular workload.
“The beauty of Supermicro is that we can provide the most advanced AI infrastructure building blocks available from NVIDIA for these modern workloads,” Srivastava says. “The entire system is plug-and-play, meaning once you receive the system, it’s ready to run your applications.”
The building-block approach is also key to future-proofing because systems can be
upgraded and reconfigured as new and better components become available or AI needs change. When a supplier introduces a nextgeneration GPU with better performance and more power, Supermicro’s existing systems can be easily upgraded with that GPU. “This is a unique aspect of our systems, that with minimal changes, you can be ready for the next generation of accelerators,” Srivastava says.
Moreover, the company provides its customers with customized service and expertise, helping them configure their systems to optimize performance and efficiency.
One large enterprise customer asked Supermicro if they could help solve a problem: It had fans running to cool the system, but one of the fans was providing more cooling than needed—wasting energy. Supermicro helped reconfigure the system into two separate cooling zones, reducing energy use in the part of the system that needed less cooling.
“We ended up saving that customer a lot of energy, and we learned things from the process we’re now applying to our next generation of systems,” Srivastava says.
Supermicro’s modular and customized approach ultimately means customers benefit from a flexible AI architecture that enables seamless scalability—allowing them to quickly ramp up their capacity and compute power as their AI needs evolve and expand. The systems are also optimized to run the applications customers use to get the most out of their AI.
“This flexibility and scalability allows our customers to get the most out of their AI workloads today while being able to expand and grow fast in the future,” he adds.
“ We ended up saving that customer a lot of energy, and we learned things from the process we’re now applying to our next generation of systems ”
In technology, Artificial Intelligence (AI) and cloud solutions are not just buzzwords – they are transformative forces reshaping industries and driving innovation.
For businesses eager to harness these powers, the journey can seem daunting. This is where Tarsus On Demand (TOD) steps in, offering a guiding hand through the AWS Partner Onboarding program and building the cloud ecosystem that South Africa’s small and medium businesses (SMBs) need.
A Partnership Built on Expertise
Imagine a world where your business operations are streamlined, customer interactions are personalised, and decisionmaking is data-driven. This is the promise of AI, and Tarsus On Demand (TOD) is here to make it a reality. With years of experience in cloud solutions, TOD has become a beacon of expertise, helping businesses navigate the complexities of AI integration.
“Our mission at Tarsus On Demand is to empower businesses with innovative
solutions that drive efficiency and growth. Partnering with AWS allows us to deliver unparalleled value to our clients, helping them stay ahead in a competitive market,” says Chantelle Trollip AWS Program Manager at Tarsus On Demand.
Building the Cloud Ecosystem
South Africa’s SMBs Need
South Africa’s SMBs have a critical role to play in the economic future, employing between 50% and 60% of the country’s workforce and acting as the engine of job growth. However, to compete in the digital economy, they need the right kind of help.
“SMBs have to equip themselves to compete in a marketplace that is increasingly digital and highly fluid. They need to digitalise, which means not only using technology smarter but also adapting their business models to take advantage of new opportunities,” says Chantelle.
The AWS Launch: A Gateway to Collaboration
Tarsus On Demand recently launched its partnership with AWS to the public, hosting events in Johannesburg and Cape Town. These events were more than just gatherings—they were a convergence of minds, ideas, and opportunities. TOD’s team of experts shared insights and explored collaboration possibilities, making it a prime opportunity for AI enthusiasts, potential AWS partners, and businesses looking to innovate.
This strategic collaboration between TOD and AWS aims to provide enhanced cloud services and support for independent software vendors (ISVs) and managed service providers (MSPs) across subSaharan Africa. The partnership is set to transform the cloud services landscape by leveraging the strengths of both TOD and AWS, empowering TOD’s partners with comprehensive multi-cloud strategies and advanced technical support, driving business growth and innovation.
Partnering with Tarsus On Demand through the AWS Partner Onboarding program is more than just a business decision—it’s a strategic move towards innovation and growth. We invite you to join us at the AI Summit, where you can learn more about our offerings, engage with our experts, and explore how we can work together to unlock the full potential of AI and cloud solutions.
In a world where technology is constantly evolving, staying ahead means embracing change and seizing opportunities. With Tarsus On Demand and AWS by your side, the future of AI and cloud solutions is within your grasp. Together, we can unlock untapped AI and cloud opportunities.
Partner with Africa's leading data science platform to access the power of 72 000+ data scientists for your business.
Partner with Africa's leading data science platform to access the power of 72 000+ data scientists for your business.
Through our partnership program, organisations can tap into this pool of talent, access machine learning and AI solutions, and stay at the forefront of innovation.
Through our partnership program, organisations can tap into this pool of talent, access machine learning and AI solutions, and stay at the forefront of innovation.
Our quick turnaround means you keep improving with the smartest minds in data science on your team.
Our quick turnaround means you keep improving with the smartest minds in data science on your team.
w ww.zindi.africa
w ww.zindi.africa
tam@zindi.africa
tam@zindi.africa
+44 7 522 120077
+44 7 522 120077
/ By Chris Szala, VP Sales, Verint South Africa /
An incredible aspect of modern technology is the helping hand it gives to humans. In most cases, technology makes our lives easier by taking care of tasks that we could be doing but would prefer to concentrate our efforts elsewhere.
Much like a GPS helps us navigate so we can focus on driving, AI (artificial intelligence) powered solutions can support teams across a contact centre, and beyond, to elevate CX while also increasing operational efficiency. The most effective AI solutions each carry out specific tasks that, in turn, increase your workforce’s capacity. While the most obvious use cases mainly involve helping your contact centre agents, AI capabilities can also transform the workloads of supervisors and data analysts –saving countless hours across your company. Read on to see where you can leverage AI to create extra capacity across the workforce.
How AI Can Create Extra Agent Capacity
Automation and Autonomy: No Need for Human Input
IVAs (Intelligent Virtual Assistants) are perhaps the best-known example of AI’s ability to take the strain from contact centre teams. In effect, IVAs act as the gatekeeper for interactions, resolving the majority of customer queries.
A well-built IVA gives customers the ability to self-serve, and ideally contains the conversation without escalation to a human agent. This can significantly reduce the volume of conversations an agent is required to handle, giving them greater capacity to handle more complex cases.
Manually writing call summaries can take up to 90 seconds out of an agent’s day per call, and you’re relying on their memory of an often complex conversation for an accurate, unbiased record. Why put that pressure on your workforce when higher precision options are available?
With accurate AI-driven transcription working in tandem with generative AI, a short and precise summary can be swiftly created, virtually eliminating after-call work. Automating this task translates into companies saving millions of dollars every year, simply by giving agents an extra 60-90 seconds back at the end of every customer call.
How AI Can Create Extra Supervisor Capacity
Efficient Staffing
Contact centre supervisors undertake a delicate balancing act; don’t overstaff when it’s quiet, but make sure enough agents are working to keep wait times down. Also, ensure a good work/life balance is possible, among all other responsibilities.
Forecasting solutions use AI to accurately predict volume and customer demand, which means that without being data specialists, contact centre managers can confidently set shift patterns that meet demand and reduce staffing costs.
The question of enabling a good work/ life balance has become more prominent since the rise of remote working. Agents are not always tied to an office, so more flexible working conditions are possible, and expected.
AI’s ability to forecast customer volume means contact centre staffing needs can also be predicted. Technology that enables agents to adjust their shifts while also imposing limits that prevent a drop in service levels is ideal for both staff and supervisors. On the one hand, agents can adjust their work to fit around their lives, and on the
other, management has more capacity to concentrate on other, more important aspects of their job. This works to minimize the chance of a staffing emergency through increased attrition or absentees, which drastically reduces staffing-related costs.
The accuracy of your engagement data dictates the quality of your insights. AI-powered transcription creates highly accurate engagement data and that data helps CX leaders to build data-driven strategies. When combined with interaction data, transcription provides analysts with access to an ever-increasing source of customer behavioral insights.
By analyzing these precise transcripts, analysts can more easily understand customer interactions by channel and any challenges they encounter. They can also determine how effectively bots and agents are resolving customer issues. These insights enable companies to build more successful CX strategies, helping to increase customer satisfaction, loyalty, and, ultimately, revenue.
Leveraging AI-powered bots across voice, digital and social channels is the key to creating extra capacity for your workforce. While AI working on the frontline of customer engagement via IVAs is the most direct way to elevate CX and improve operational efficiency, there are plenty of other use cases which are just as effective. Whether your AIpowered solutions are helping with contact centre scheduling or unearthing insights that contribute to data-driven strategies, each creates an environment that continually optimizes your teams’ output.
At Verint, openness is not a feature, it’s a strategy. This gives you the freedom to choose the right path for your contact centre now and in the future so you can elevate customer experience and evolve into best-in-class operations.
Open CCaaS delivers unmatched freedom and flexibility to drive CX automation. Leverage data, AI and bots to elevate your CX and lower cost.
Verint understands your CX Automation priorities extend far beyond telephony. With our next-gen Open CCaaS platform, you have open data, open AI, open best-of-breed solutions, open ecosystems, and an open enterprise. Everything is open.
South Africa’s business landscape is becoming increasingly challenging, plagued by currency volatility, payment delays, pervasive corruption, and an escalating risk of invoice fraud. With no signs of these issues slowing down, the question arises: how can businesses defend against these escalating threats? What are the most effective strategies for safeguarding against fraud? And how can advanced technology like Artificial Intelligence (AI) be harnessed in this fight?
While your business may not control currency fluctuations or corruption, you can actively address payment delays and invoice fraud. The Accounts Payable (AP) department, which manages an organization’s cash flow via the invoice workflow and protects sensitive data, plays an obviously critical role in your financial operations. However, this essential role makes them a prime target for cyberattacks.
In fact, invoice fraud has rapidly evolved from a potential risk to a daily threat for South African businesses of all sizes and industries. The frequency and severity of these cyber-attacks are becoming more sophisticated, with criminals increasingly impersonating legitimate suppliers to divert payments into fraudulent accounts or manipulate people into disclosing confidential information. This alarming trend is so great that it has prompted institutions like the World Bank to issue global warnings, underscoring the urgent need for business to fortify their defenses.
In response to these escalating risks, AI is emerging as an indispensable tool and vital ally. This advanced technology enables realtime detection of anomalies by analyzing vast amounts of transaction data to spot
suspicious patterns. AI also automates verification processes, significantly reducing the chance of human error – all within minutes.
However, AI doesn’t stop there. Beyond detection, it plays a proactive role, anticipating potential threats and providing a strong, adaptive defense against fraud attempts.
Yooz:
Yooz identified the potential of AI early on, seamlessly integrating it with cutting-edge technology to develop a comprehensive accounts payable automation solution. The goal? To safeguard your company’s financial health and reputation from the very start of the invoice process.
By leveraging advanced AI, Yooz delivers a comprehensive suite of features that are designed to safeguard your operations against fraud. This includes real-time anomaly detection, automated verification processes, secure document approvals, and more. All to anticipate potential threats, reinforce your defenses, and ensure strong data protection.
To stay ahead in the ongoing battle against fraud and elevate your financial operations, visit our website and request a personalized demo today! https://interfacesystems. co.za/ solutions/yooz/ or email: info@Interfacesystems. co.za
/ Read original article here /
Artificial Intelligence (AI) is more than a technological leap; it’s a transformative force reshaping our world. With far-reaching impacts across economics, society, and geopolitics, AI is driving revolutionary changes in healthcare, agriculture, finance, and education.
For Africa, AI is a strategic asset pivotal to achieving the aspirations of Agenda 2063 and the Sustainable Development Goals (SDGs). It promises to ignite new industries, fuel innovation, and create high-value jobs while preserving and advancing African culture and integration.
In a landmark decision, the African Union Executive Council endorsed the Continental AI Strategy during its 45th Ordinary Session in Accra, Ghana, on July 18-19, 2024. This strategy underscores Africa’s commitment to an Africa-centric, development-focused approach to AI, promoting ethical, responsible, and equitable practices.
The Continental AI Strategy calls for unified national approaches among AU Member States to navigate the complexities of AI-driven change, aiming to strengthen regional and global cooperation and position Africa as a leader in inclusive and responsible AI development.
Access the Strategy: [English] [French] [Arabic] [Portuguese]
Related Resources:
African Union Digital Transformation Strategy
African Child Online Safety and Empowerment Policy
African Union Data Policy Framework
African Digital Compact
Harnessing AI for Africa’s Development and Prosperity
/ By Gregory Williams, RVP South Africa /
Imagine your business processes not only running themselves but actively learning, adapting, and improving without constant human intervention. This is not science fiction – it is the promise of agentic AI, and it is closer than you might think.
While agentic AI is an emerging technology, it will be propelled by the foundations of robotic process automation (RPA) and AI.
Defining agentic AI
Agentic AI is the next step in the evolution of RPA, automation, and AI. It can be defined as AI systems capable of controlling and managing the flow of business processes with increasing levels of independence and intelligence.
Excitement about the potential of AI in the enterprise remains high. The problem is that there is a gap between the promise of AI and the reality of business outcomes being achieved through AI today. Aiming to bridge this gap, enterprises are seeking the operating model of the future to truly create a new way of working.
AI uses the basic building blocks of automation. With RPA, people create and maintain automations supported by an automation platform. As the evolution of human and AI collaboration evolves across business processes, Agentic AI will be capable of assuming higher levels of intelligence.
This new way of working goes beyond traditional automation, where humans create and maintain automations; it leverages AI agents that can learn, adapt, and even make decisions within certain limits.
Integrating RPA and agentic AI, also known as agentic process automation (APA), presents a future where automation encompasses both routine operational efficiencies and advanced cognitive functions, giving organizations the tools to transform both their workflows and decision-making processes. This approach not only increases the scope of automation but also enhances the ability of businesses to respond to complex and changing environments with greater agility.
The building blocks of agentic AI Agentic AI does not emerge in isolation but complements and builds upon the foundation of RPA and other AI technologies that businesses may already be using. Companies’ current investments in automation are safe and will still be important parts of this new era of AI-driven processes.
For regular operations, expensive AI implementations may not be necessary. Enterprises will not need costly large language models (LLMs) to support lowervalue task automation. This approach ensures that businesses can costeffectively grow their AI capabilities.
Who can benefit from agentic AI?
Every industry can benefit, from financial services and healthcare as well as manufacturing and retail. In Africa, where businesses increasingly are using cuttingedge technology platforms to solve problems and be more competitive, agentic AI will be a substantial leap forward. From automating mundane tasks to navigating ambiguity and uncertainty, the synergy
between RPA and Agentic AI has the ability to revolutionize the way businesses operate.
Small and medium businesses (SMEs) can also greatly benefit from agentic AI as well, by automating complex processes across various business units as is taking place within enterprises.
Implementing agentic AI in African businesses
For African companies looking to embrace agentic AI, here are some key considerations:
Start with a clear plan: Look at your current processes and find areas where agentic AI could make the most difference.
Build on existing automation: If you are already using RPA or other automation tools, you are in a great position to use agentic AI.
Focus on data quality: Ensure your data infrastructure is robust and your data is clean and well-organized.
invest in skills development: Invest in training your workforce to work alongside and manage AI systems.
Start small and scale: Begin with pilot projects in non-critical areas of your business.
What businesses should know
Agentic AI is exciting, but it is important to know what it can and cannot do.
it is an evolution, not a revolution: Agentic AI builds on existing technologies. It is not about replacing current systems but improving them.
Human oversight remains crucial: Human judgment and oversight remain essential, especially for critical decisions.
Ethical considerations are important: Businesses must be careful about how AI decisions affect their
...continues on page 49
/ By Junaid Kleinschmidt and Lutendo Munzhedzi – Accenture /
The South African retail landscape is contending with a range of challenges, such as high unemployment, severe inflation, “shrinkflation”, and the lingering effects of Covid-19. These pressures have transformed consumer expectations and added layers of complexity to their shopping decisions. Retailers must now make faster and more efficient decisions to meet these evolving demands. With advanced technology at our disposal, how can Gen AI ease the decision-making process for consumers?
The unemployment rate in South Africa has reached concerning levels, standing at approximately 33%, reported by Stats SA as of Q1 2024. This widespread unemployment has led to reduced financial freedom for many households, resulting in diminished disposable income. Additionally, inflation has eroded consumers' spending power. Consumer spending in South Africa is said to have declined from the fourth quarter of 2023 to the first quarter of 2024, as an effect of significant constraints on disposable income and overall household finances.
An added issue is “Shrinkflation” - the practice of manufacturers, and therefore retailers, reducing the mass / weight of products, but maintaining the same or a higher price point. This forces consumers to pay increasing prices for diminishing returns. According to the annual Accenture Life Trends Report, the tightening cost of living has resulted in the consumer’s heightened
attention to value, resulting in lower susceptibility to settle for less.
COVID-19 has had a profound impact on trading activity and the overall financial wellbeing of consumers affecting how and what they buy. The pandemic has fundamentally reshaped consumer engagement, whilst also driving a hybrid shopping model where South Africans research products online, but still explore and purchase in-store, emphasising the need for tools that shape the customers experience from an omni-channel perspective. This calls for strategy adaptations to accommodate for changing consumer needs
The overwhelming decision-making task of what to buy, when to buy it, and where to buy it are being dealt with by consumers as they seek to accommodate their needs with limited resources. Rising prices have forced many to consider alternative options to the products they were previously accustomed to, seeking the right balance between quality and cost. South African consumers now
prioritise value for money, requiring them to evaluate product information to determine if these alternatives still meet their needs. The hybrid nature of modern shopping introduces more layers, as they must now navigate multiple channels - both online and in-storethroughout their shopping journey. The sheer volume of choices can make shopping an exhausting task, rather than an enjoyable one.
A recent Accenture study has reflected that 75% of retail executives view Gen AI as core to their business revenue growth. As South Africans pay higher attention to value, decision-making across a range of products with added cost-consciousness places a heavy burden on the consumer, solidifying the need for more effective tools and strategies that can simplify and optimise this process for consumers. According to the Accenture Life Trends Report, there is a large opportunity space to drive targeted customer interaction, engagement, and personalisation. With South Africa’s diverse demographic, the challenges faced require dedicated capabilities to address these issues, and Gen AI can provide a huge competitive advantage. Accenture reports that 72% of retailers are planning to leverage Gen AI for operation reinvention - and capitalising on these opportunities can offer great rewards for both consumers and retailers, as seen in the following examples:
Gen AI can be used to augment existing capabilities and enhance the value-add for both the consumer and retailer, examples of this include conversational AI capabilities such as chatbots and virtual shopping assistants for advice and recommendations. With Gen AI’s power of creation through prompts, this can generate customised product baskets based on consumer preferences and budget constraints. This is crucial for shoppers seeking smaller yet value-packed baskets that do not deviate too much from what they would typically buy, assisting consumers with limited budget constraints on decision-making for product selection.
When it comes to personalised shopping assistance and tailored product baskets, Walmart is a great example of this, having integrated Gen AI into their search function, customers are now able to input a prompt and receive tailored responses, including the ability to offer recommendations on an optimal basket of goods based on needs and constraints.
The development of Gen AI tools that can create product descriptions, offers
...continues on page 46
/ Read original article here /
Food security, healthcare and exploring the cosmos are among the ways students of a new panAfrican Master’s program aspire to apply AI
At Google DeepMind, we’re committed to supporting the next generation of artificial intelligence (AI) leaders to help build a stronger, more diverse and inclusive global AI community. This includes increasing access to AI and science through education.
Last year, we partnered with the African Institute for Mathematical Sciences (AIMS), Africa’s first network of centers of excellence in mathematical sciences, to launch an AI
for Science Master’s program, with a $4.5M grant from Google DeepMind.
This funding helps AIMS provide full scholarships, equipment and compute to talented local students, giving them access to advanced studies in mathematics, AI and machine learning from world-class academics at AIMS South Africa. Students have the opportunity to accelerate scientific discovery, with mentoring and support from Google DeepMind’s researchers and engineers.
This summer, the first cohort of students graduated at a ceremony at the AIMS campus in Cape Town, South Africa. As the next generation of AI leaders in Africa, Béria Chingnabé Kalpélbé, Olivier Mahumawon Adjagba and Diffo Mboudjiho Annette Dariose shared their experiences in pioneering AI research and what they’re hoping to achieve with their work.
...continues on page 46
...continued from page 44
consumers quick relevant insights into the best product options based on their needs. This makes it easier to identify the most cost-effective products, and eliminates the added responsibility of having to sift through copious amounts of information on products.
3. Dynamic Promotion Planning & Optimised Communication and Timing
Classical AI, when combined with Gen AI, can play a key role by developing models and algorithms for dynamic promotion planning, allowing retailers to adjust promotions based on real-time data, inventory levels, and consumer behaviour patterns. To prevent marketing fatigue, introducing solutions like contact frequency models that predict optimal frequency and the preferred customer contact platform, with discounted price alerts, based on data from customer interactions, purchase history, segmentation, and product affinity, can aid in ensuring customers are not overwhelmed, while maintaining a positive brand relationship.
There is immense power to be unlocked from Gen AI - Shopify for example, has incorporated multiple Gen AI solutions into their products. Since 2023, the commerce platform has introduced an AI-generated product description capability to send emails at optimal times, and to tailor communications for timelier promotions and email correspondence.
Gen AI offers promising capabilities to retailers to increase their competitive edge and keep them ahead of evolving demands. For consumers, it offers to decrease the load on decision-making, whilst accommodating for the nuances that exist in different communities, and creating a better consumer experience.
Time for Advancement is Now
We are on the verge of a major shift, with innovation moving to implementation faster than ever, giving the market a new edge. As South Africa continues to embrace digital transformation, Gen AI has the potential to revolutionise the retail experience. By offering personalised, efficient, and culturally relevant shopping experiences, Gen AI can help alleviate the overwhelming choices consumers face, making shopping more enjoyable and accessible for all.
For South African retailers, the message is clear: Those who harness the power of Gen AI will be better positioned to meet the evolving needs of their customers, driving loyalty and growth in a competitive market.
...continued from page 45
Béria: Innovating for better food security
Sustainability is a top priority for Béria, originally from Chad. “I hope to develop solutions for sustainable agricultural development that will benefit both people and the planet by integrating principles of renewable energy, precision farming, and ecological preservation in my work,” he says.
“Beyond agriculture, AI offers significant potential to enhance the resilience of Africa's natural environments,” Béria adds. “By implementing AI-powered monitoring and decision-support systems, we can safeguard Africa's precious green areas and biodiversity for future generations.”
Olivier: Pioneering virus transmission research through the lens of climate change
Olivier’s passion for applying mathematics to complex problems led him to AIMS South Africa: “Throughout my academic journey, I’ve been fascinated by the power of mathematics, particularly in addressing real-world challenges through AI,” he says. “A solid foundation in mathematical sciences is crucial for driving progress in areas such as healthcare, climate science and technology — and I’m eager to be at the forefront of these advancements.”
Originally from Benin, Olivier now looks to apply this approach to data from African countries to help understand the spread of dengue fever. “Using advanced AI techniques, I hope to create more accurate prediction models to inform public health strategies and interventions, ultimately contributing to the control and prevention of this viral disease.”
Discussing the personal impact of his scholarship, Olivier recounts, “Without it, pursuing advanced studies at such a prestigious institution would have been financially unattainable for me. This support enabled me to fully immerse myself in AIMS' rigorous academic environment, so I could engage deeply in coursework, collaborate with professors and peers, and contribute meaningfully to research projects.”
Diffo: Unraveling the secrets of our universe
Diffo, from Cameroon, is fascinated by the big questions beyond Earth — which is what drew her to the Square Kilometre Array (SKA), the largest and most sensitive radio telescope on the planet.
“Understanding the 21cm line provides insights into the early universe, the formation of the first stars and galaxies, and the structure of the
cosmos,” Diffo explains. “By applying Markov chain Monte Carlo (MCMC) techniques, I hope to improve the accuracy and efficiency of extracting these faint signals from SKA data, potentially leading to more precise cosmological models and a deeper understanding of the future evolution of the universe.”
For those considering similar studies, Diffo offers a few words of advice: “Stay curious, be persistent and embrace interdisciplinary learning. Engaging in hands-on projects, collaborating with peers, and seeking mentorship from AI experts can greatly benefit your learning experience and career prospects.”
This work builds on our existing commitments in the region, including our support of the Deep Learning Indaba through volunteering and funding since its inception in 2017, the recent launch of our Experience AI education program across Africa, which has already engaged local educators working with more than 30,000 young people, and additional educational funding, which has been used by three further African universities to offer a total of over 40 postgraduate scholarships since 2020. Increasing representation in the field of AI research offers a much-needed opportunity to bring diverse values, perspectives, and concerns into conversations about the design and deployment of this transformative technology. We hope our support for AIMS not only serves to build a more global and inclusive AI ecosystem, but also helps students make new scientific discoveries that benefit their local communities and the entire globe.
With special thanks to Ulrich Paquet, research scientist at Google DeepMind, who is serving as executive director at AIMS South Africa until 2027 and helped launch the AI for Science Masters program. Paquet continues to hold a dual affiliation with Google DeepMind.
We would also like to thank the University of Cape Town, Stellenbosch University, and the University of the Western Cape, the degree-granting institutions underpinning this program at AIMS South Africa, for their long standing academic support.
/ Read original article here /
Mistral-NeMo-Minitron 8B is a miniaturized version of the recently released Mistral NeMo 12B model, delivering high accuracy combined with the compute efficiency to run the model across GPUaccelerated data centers, clouds and workstations.
Developers of generative AI typically face a tradeoff between model size and accuracy. But a new language model released by NVIDIA delivers the best of both, providing state-of-the-art accuracy in a compact form factor.
Mistral-NeMo-Minitron 8B — a miniaturized version of the open Mistral NeMo 12B model released by Mistral AI and NVIDIA last month — is small enough to run on an NVIDIA RTX-powered workstation while still excelling across multiple benchmarks for AI-powered chatbots, virtual assistants, content generators and educational tools.
Minitron models are distilled by NVIDIA using NVIDIA NeMo, an end-to-end platform for developing custom generative AI.
“We combined two different AI optimization methods — pruning to shrink Mistral NeMo’s 12 billion parameters into 8 billion, and distillation to improve accuracy,” said Bryan Catanzaro, vice president of applied deep learning research at NVIDIA.
“By doing so, Mistral-NeMo-Minitron 8B delivers comparable accuracy to the original model at lower computational cost.”
...continues on page 49
...continued from page 47
“ We combined two different AI optimization methods — pruning to shrink Mistral NeMo’s 12 billion parameters into 8 billion, and distillation to improve accuracy,” said Bryan Catanzaro, vice president of applied deep learning research at NVIDIA. ”
Unlike their larger counterparts, small language models can run in real time on workstations and laptops. This makes it easier for organizations with limited resources to deploy generative AI capabilities across their infrastructure while optimizing for cost, operational efficiency and energy use. Running language models locally on edge devices also delivers security benefits, since data doesn’t need to be passed to a server from an edge device.
Developers can get started with MistralNeMo-Minitron 8B packaged as an NVIDIA NIM microservice with a standard application programming interface (API) — or they can download the model from Hugging Face. A downloadable NVIDIA NIM, which can be deployed on any GPU-accelerated system in minutes, will be available soon.
For a model of its size, Mistral-NeMoMinitron 8B leads on nine popular benchmarks for language models. These benchmarks cover a variety of tasks including language understanding, common sense reasoning, mathematical reasoning, summarization, coding and ability to generate truthful answers.
Packaged as an NVIDIA NIM microservice, the model is optimized for low latency, which means faster responses for users, and high throughput, which corresponds to higher computational efficiency in production.
In some cases, developers may want an even smaller version of the model to run on a smartphone or an embedded device like a robot. To do so, they can download the 8-billion-parameter model and, using NVIDIA AI Foundry, prune and distill it into a smaller, optimized neural network customized for enterprise-specific applications.
The AI Foundry platform and service offers developers a full-stack solution for creating a customized foundation model packaged as a NIM microservice. It includes popular foundation models, the NVIDIA NeMo platform and dedicated capacity on NVIDIA DGX Cloud. Developers using NVIDIA AI Foundry can also access NVIDIA AI Enterprise, a software platform that provides security, stability and support for production deployments.
Since the original Mistral-NeMo-Minitron 8B model starts with a baseline of state-of-theart accuracy, versions downsized using AI Foundry would still offer users high accuracy with a fraction of the training data and compute infrastructure.
To achieve high accuracy with a smaller model, the team used a process that combines pruning and distillation. Pruning downsizes a neural network by removing model weights that contribute the least to accuracy. During distillation, the team retrained this pruned model on a small dataset to significantly boost accuracy, which had decreased through the pruning process.
The end result is a smaller, more efficient model with the predictive accuracy of its larger counterpart.
This technique means that a fraction of the original dataset is required to train each additional model within a family of related models, saving up to 40x the compute cost when pruning and distilling a larger model compared to training a smaller model from scratch.
Read the NVIDIA Technical Blog and a technical report for details.
NVIDIA also announced this week NemotronMini-4B-Instruct, another small language model optimized for low memory usage and faster response times on NVIDIA GeForce RTX AI PCs and laptops. The model is available as an NVIDIA NIM microservice for cloud and on-device deployment and is part of NVIDIA ACE, a suite of digital human technologies that provide speech, intelligence and animation powered by generative AI.
Experience both models as NIM microservices from a browser or an API at ai.nvidia.com.
See notice regarding software product information.
...continued from page 42
values, data privacy, and possible biases in AI decision making.
it is about augmentation, not replacement: Agentic AI is about improving human skills, not replacing them. It does this by helping employees focus on more important and creative tasks.
Continuous learning is key: Both the AI systems and the people working with them will need to change and learn as the technology changes.
The future is both robotic and agentic
As we look ahead, we know that the businesses to succeed will be those that successfully combine human creativity with AI capabilities. Agentic AI is a big step forward in this integration. It gives African businesses new ways to think, compete, and grow in a world where everything is digital.
With many businesses already using AI, it is important to know that it is not a one-size-fits-all solution. Instead, it is a fabric that can create solutions that are specific to an organization’s entire operating model, and businesses will need to continue to tailor their AI implementations to their unique needs and challenges.
For African enterprises, from startups in Stellenbosch to established corporations in Johannesburg, agentic AI presents an opportunity to leapfrog traditional development stages. By embracing this technology, they can enhance efficiency, improve decision making, and deliver superior customer experiences. However, the key to success lies not just in adopting the technology, but in reimagining business processes with a human-AI collaborative approach at their core.
Google Translate is using artificial intelligence (AI) to add 110 new languages to its translation service and about a quarter of the new languages come from Africa. The new additions represent Google Translate's largest expansion of African languages to date.
Google has committed to build AI models that will support the 1,000 most spoken languages around the world.
(Source: DC Studio on Freepik)
/ Read original article here /
The new list includes the languages of Fon (primarily spoken in Benin); Kikongo (Republic of Congo, Gabon and Angola); Luo (Kenya and Tanzania); Ga (South Eastern Ghana); Wolof (Senegal); Siswati (Eswatini and South Africa); Tshivenda (South Africa, Zimbabwe and Mozambique); Afar (Djibouti, Eritrea and Ethiopia); NKo (Guinea, Guinea-Bissau, Mali, Ivory Coast, Burkina Faso and some other West African countries); Tamazight (also known as Amazigh and spoken in Morocco and Algeria).
Google said the new list of 110 new languages represents more than 614 million speakers, opening up translations for around 8% of the world's population.
"Some are major world languages with over 100 million speakers. Others are spoken by small communities of Indigenous people, and a few have almost no native speakers but active revitalization efforts," the company said.
AI models assist with translation
The additions are part of Google's 1,000 Languages Initiative – a commitment to build AI models that will support the 1,000 most spoken languages around the world –and the latest expansion uses the PaLM 2 large language model
PaLM 2 is Google's next-generation language model with improved multilingual, reasoning and coding capabilities "that builds on Google's legacy of breakthrough research in machine learning and responsible AI," the company said.
In May 2022, Google Translate also added 24 more languages using Zero-Shot Machine Translation, where a machine learning model learns to translate into another language without ever seeing an example.
The 2022 update added the African languages of Bambara (Mali); Ewe (Ghana and Togo); Krio (Sierra Leone); Lingala (Democratic Republic of the Congo, Republic of the Congo, Central African Republic, Angola and South Sudan); Luganda (Uganda and Rwanda); Oromo (Ethiopia and Kenya); Sepedi (South Africa and Botswana); Tigrinya (Eritrea and Ethiopia); Tsonga (Eswatini, Mozambique, South Africa and Zimbabwe); and Twi (Ghana).
"Google Translate breaks down language barriers to help people connect and better understand the world around them," the company said in a statement.
For South Africa specifically, the addition of Siswati and Tshivenda brings the total number of South African languages on Google Translate up to nine out of 12 (with South African Sign Language [SASL] added to the previous 11 official languages in July 2023).
"Language is an essential part of how people communicate, understand, and interact with the world around them so we are very excited about this latest expansion," Google South Africa's communications manager, Siya Madikane, told Connecting Africa via email.
Madikane said a lot of consideration goes into new language additions for Google Translate, ranging from which languages to include to the use of specific spellings.
"Many languages do not have a single, standard form, so learning the specific dialect that is spoken the most in an area is more feasible. Our approach has been to prioritise the most commonly used varieties of each language," he added.
Google said that PaLM 2 played a crucial role in assisting the team translate more efficiently and learn languages that are closely related to each other.
"As advancements in technology continue to occur, Google will continue to partner with expert linguists and native speakers – to support even more language varieties and spelling conventions over time," Google added.
/ Read original article here /
Amid the rising threat of severe weather patterns, three organisations are collaborating to launch a social coding challenge to develop impactful artificial intelligence (AI)-driven solutions that address this issue.
The International Research Centre on Artificial Intelligence (IRCAI), an organisation under the auspices of Unesco; data science platform Zindi; and Amazon Web Services (AWS) have launched the “AI for Equity Challenge: Climate Action, Gender and Health”.
The six-month global virtual contest focuses on creating solutions that tackle gender and health equity issues exacerbated by climate change, says a statement.
It will support researchers, non-profits and start-ups applying advanced cloud computing and AI solutions that address some of the UN Sustainable Development Goals (SDGs).
The challenge will focus on building new programming that supports more equitable futures at the intersection of gender, health and climate action, targeting SDGs three (good health and well-being), five (gender equality) and 13 (climate action).
Davor Orlic, COO of IRCAI, says: “As artificial intelligence rapidly evolves, it is crucial that we harness its potential to address real-world challenges. At IRCAI, our mission is to guide the ethical development of AI technologies, ensuring they serve the greater good and are inclusive of marginalised AI communities.
“This challenge, in collaboration with AWS and Zindi, is an opportunity to discover and support the most innovative minds that are using AI and advanced computing to create impactful solutions for the climate crisis.”
According to the statement, participants will use data made available through the Amazon Sustainability Data Initiative, with a preference towards building on AWS AI services and infrastructure, to develop solutions.
A steering committee made up of members of Naixus, a network of AI and sustainable development research organisations, being convened through IRCAI, will select proposals that demonstrate the most innovative use of advanced AI technologies, and the biggest promise for global impact on health and gender equity, and climate action.
Selected teams will receive expert guidance in AI sustainability and responsible AI, along with AWS credits to build their solutions, says the statement.
Dave Levy, vice-president of Worldwide Public Sector at AWS, adds: “Climate change is one of the greatest threats of our time, and we believe innovation is key to overcoming it.
“The AI for Equity Challenge invites innovators to bring forward their most visionary ideas, and we’ll support them with AWS resources – whether that’s computing power or advanced cloud technologies – to turn those ideas into reality.”
The statement indicates the challenge will proceed in phases, with the first round of technical proposal submissions closing on 4 November, culminating with an announcement of the winning solutions in February 2025.
To apply for the AI for Equity Challenge, click here
South Africa faces a significant education crisis, with dropout rates that undermine the country’s potential for social and economic growth. South Africa has a rising school dropout rate, with some estimates suggesting that around 60% of grade ones
In the context of Sub-Saharan Africa, it is alarrming to note the following number of children have dropped out of school
●over 20% for children aged 6–11
●over 33.3% for children aged 12–14
almost 60% for children aged 15–17
The dropout rate for higher education in South Africa is sitting at approximately 52%, and only one in three students graduates within four years. Shockingly, with all these educational factors contributing to economic stagnation, South Africa still lacks a strategic plan towards an education system aimed at discovering why we have such a high failure and dropout rate.
While remedial education could be a solution for some of these challenges, in an economy where approximately 55.5% of the population is living in poverty, access to remedial education is a luxury. At government schools where remedial education is available, the quality is higher in metropolitan areas and lower in schools situated in rural areas. Remedial education is aimed at students who have an average (or even a higher-than-average) intellectual ability but are struggling to perform well in a traditional school environment.
/ By Cleo Gwenar /
Artificial Intelligence (AI) is here, and in Africa there are a lot of opportunities and possible threats around its use. Because Africa is a developing continent, with some areas more advanced than others, it is important to understand the technological landscape socially, economically, and politically. With these considerations, AI can be effectively leveraged to positively benefit the communities it is being implemented in.
For example, in the education sector there are different types of schooling: mainstream, remedial, and special school systems. It is important to note that there is a gap in the support for remedial and special schools in terms of ensuring that children at these schools continue to receive the support they need to be successful throughout their student careers. AI technologies can be used more effectively in this sector by ensuring the children are taught using AI technologies that accommodate their physical or mental needs. This would assist a child with ADD, ADHD, dyslexia, or other types of neurodivergence by providing personalized educational material that accommodates the way students consume and interpret the content.
This can aid the learner from an early age after being diagnosed to understand how they would navigate their schooling years and help them prepare for the future more intentionally with AI as their aid as they continue their academic lives. This would aid the learners greatly and ensure that they are more comfortable learning even though they face disabilities that would have previously inhibited them from traditional learning. AI can be an enabler for improved learning experiences for the differently abled and accommodate their learning styles and abilities as well as prescribe learning models suited for each learner. This is but just one example of how AI can be utilized more, and there are many more use cases for AI in other government and private sectors.
WOMENINAI (WAI) is a non-profit dotank working towards inclusive AI that benefits global society. The organization is a community-driven initiative bringing empowerment, knowledge, and active collaboration via education, research, events, and more. Women and minorities are encouraged to become, among other things, change makers through ethical applications and responsible use of artificial intelligence.
WAI was founded in 2016 in Paris, France, by Dr. Hanan Salam, Caroline Liar and Moojan Asghari. It started out as a Facebook group that has grown into a powerful global community with more than 8,000 members and growing in 140 countries. WAI in Europe, the Middle East, and Africa (EMEA) focuses on driving innovation, partnerships, and education to address challenges facing communities.
The mission of WAI is to create a community that will work towards shaping inclusive AI for our common future. These initiatives are done through volunteering, donations, and our wonderful global partners.
These initiatives vary according to demand, and this is where WAI would be ideal in working with the educational sector, or any other sector, to find synergies and meaningful ways to get involved in seeking solutions on how AI can be incorporated in the differently abled community from early years right until post-tertiary stages.
However, this will require the government, industry experts through sponsors, members from different industries, and subject matter experts to move in the same direction in seeking solutions to ensure that the future becomes brighter for those with different abilities for the betterment of our society.
/ By Yattish Ramhorry, Chief Technology Officer, Virtual Helper /
Language models like ChatGPT have changed how we interact with technology, but there’s a big issue—many African languages are left behind. While languages like English and Chinese get plenty of resources and investment, African languages, spoken by millions, often don’t. This leaves a gap in how technology serves people across the continent.
Inclusive AI can close this gap. When AI understands and uses African languages, it becomes a tool for preserving culture, opening opportunities, and giving people access to information in their own words. It’s not just about fairness; it’s about making sure technology works for everyone.
Why African Language
Africa’s languages are diverse and beautiful, but they’re rarely represented in AI. Most models focus on global languages, like English, and Chinese, excluding millions of people. Language models for isiXhosa, isiZulu, or Swahili could change that by improving education, preserving endangered dialects, and boosting industries like healthcare and banking.
When AI includes African languages, it connects with more people, making technology more accessible and meaningful. It’s not just about building better AI—it’s about celebrating and protecting Africa’s linguistic and cultural heritage.
Spotlight on Xhosa-ZuluLlama3 Xhosa-ZuluLlama3 caught my attention while I was doing research for a recent project, and it’s truly exciting for both Africa, and South Africa. This model works with isiXhosa and isiZulu, two widely spoken languages, and it opens a whole new world of new possibilities for these communities.
This language model was created by Jacaranda Health and AI innovators. Known for their work in maternal health, Jacaranda developed this model to improve communication and accessibility in local languages.
This model is a game-changer. It’s helping people access healthcare advice, educational content, and even customer support in their own languages. And it doesn’t stop there—Jacaranda is keeping things open-source, encouraging others to adapt the model for uses like education and e-commerce.
To me, this feels like more than just progress; it’s personal. For once, AI reflects the identities and voices of Africa, and South Africa. It’s proof that when technology is built for everyone, it can transform lives.
Here’s hoping this is just the beginning.
If you're curious about exploring XhosaZuluLlama3 or want to give it a go in your own projects, you're in luck—this model is open-source, which means anyone can get their hands on it. You’ll find it on platforms like Hugging Face, a go-to spot for hosting cuttingedge language models.
Hugging Face isn’t just a place to find models; it's a space where developers collaborate, experiment, and tinker. You'll be able to find Xhosa-ZuluLlama3 there, along with some handy documentation to get you started on your own AI journey.
In addition, Jacaranda Health, the organization behind the development of this model, offers resources on how to use it, as well as detailed information about its applications in maternal healthcare and other sectors. You can visit Jacaranda Health’s website or their GitHub page to find out more and even contribute to the project if you're keen on supporting AI that serves underserved communities.
For those who prefer a more hands-on approach, you might also find integrations of Xhosa-ZuluLlama3 in various local platforms aiming to bridge the digital divide. Keep an eye on local AI communities, tech forums, and open-source projects, as many developers are working to incorporate this tool into services like customer support, education, and healthcare.
Building these kinds of models requires collaboration—lots of it. It’s not just about coding and algorithms, but about bringing together linguists, AI researchers, tech companies, NGOs, and local communities. Everyone brings something crucial to the table, ensuring that these models are not only accurate, but also culturally relevant and effective.
The beauty of open-source models like Xhosa-ZuluLlama3 is how collaboration opens up the floodgates to wider access and adoption. When these tools are made available for use across sectors like education, healthcare, and customer service, they can have a far-reaching impact. It’s about more than just one community benefiting—it’s about many, from all walks of life.
Dawood Patel, Helm CEO
The hype and talk about LLMs and Generative AI have created much anticipation in the business world around the benefits these developments can bring to both top and bottom lines.
The challenge though, is that like any hype cycle it oversimplifies the inherent nuances and fragilities of implementing Gen AI within a business – such as the application, process, and channels, to name a few.
With Helm having been at the forefront of Natural Language Processing developments for the last eight years, we have learnt a great deal - we have the scars to prove it, and the experience that goes with it. Coupled with the rapid advancements in the Generative AI space, we believe we have now identified the safest way to implement a GenAI solution within a business environment that allows organisations to realise the benefits they’ve envisaged.
As many companies have found out the hard way – it's not as simple as just plugging ChatGPT or Llama 3 into your systems or processes. We also need to consider whether ‘traditional’ machine learning or well thought-out UX will do the trick. Or perhaps how we combine all three.
Like any tech implementation, defining a problem statement is critical, as is assessing whether the tech you intend to use is actually going to solve the problem. We have seen many organisations excitedly implementing popular solutions only to rue the haste of their decision further down the line.
In situations such as the one we find ourselves in now, the confluence of service design and data science capabilities come to the fore - skills we have been honing and mastering since 2016. I cannot stress this enough – it is crucial to understanding brand and technology landscapes before delving into the intricacies of LLMs and Generative AI. In fact, it’s so critical that we won’t start development on any such projects until we’re satisfied that these factors have been considered. Only then can we, with a clear conscience, begin to build a GenAI solution.
Once that hurdle has been successfully negotiated, one needs to understand how LLMs function in the Gen AI domain – that is the key to unlocking their potential. In order to simplify the understanding of how a LLM needs to be structured for an implementation, we need to deconstruct its building blocks:
A language model generally functions for a single language - English, French, Sotho, Arabic, Swahili etcetera. Unfortunately, not all languages are the same - the vast majority are underresourced and therefore require specialist training in order to facilitate human/ machine interaction. A well-resourced language typically has to process at least 10,000 books in order for the corpus of data to be useful. As Helm we have designed models that work very well under the low-resource constraints. The language model naturally has access to data in order to make it relevant to the application - this is the basis of LLMs such as GPT4, Claude, Gemini, Llama 3 etc - we have integrated these models where necessary to deliver some of our applications.
Once you have a baseline language model, it then needs to be tuned for domain-specific interactions. For example, the word ‘balance’ has different connotations within the finance realm as opposed to telecommunications – one provides a view on your outstanding credit card or investment account balance, whereas a telco would be referring to data bundles, minutes and so on. This requires training by highly educated data scientists.
...continues on page 60
...continues on page 57
In a rapidly evolving technological landscape, access to quality education in Science, Technology, Engineering, and Mathematics (STEM) is critical for empowering learners in under-resourced schools and communities. United Siyafunda NPC has been actively working to foster this access, particularly in township schools.
Founded in 2020, the organization aims to promote access to STEM education in under-resourced schools and communities through the provision of coding and robotics training programs, and recently – through NMB TECH, the private leg of the hybrid structured social enterprise - by developing affordable EdTech resources for marginalized communities.
At the helm of United Siyafunda NPC are Veli Mokhabodi, the Managing Director, Dieketseng Nibe, the Operations Director, and Musa Baloyi, the Technical Director. Their leadership has been instrumental in expanding the organization's reach and impact. Under this leadership, United Siyafunda NPC was established in 5 Communities in Gauteng, South Africa, providing thousands of learners with access to coding, robotics, and digital training programs.
Recent accomplishments include winning the 2023 African Impact Challenge: Community Impact Stream for work
“ Looking ahead, United Siyafunda NPC aims to expand its reach further into rural areas where access to technology is even more limited. The organization's vision includes not only increasing its impact but also erecting community digital hubs while working to enable access to EdTech resources in under-resourced and marginalized communities and schools. ”
done in township schools and robotic clubs. The organization partnered with local stakeholders in governments and educational institutions in the Kingdom of Lesotho to facilitate the 2024 Lesotho Robotics Expo, which has promoted, showcased, and enabled the implementation of innovative Coding and Robotics training programs and contests.
To combat challenges faced by underresourced schools and communities, concerning the high cost of robotics equipment, United Siyafunda NPC is actively developing solutions aimed at reducing costs associated with coding and robotics education.
Starting with the development of a zerorated Learning Management System (LMS), running on an AI platform that leverages neuroscience and quality STEM content to deliver a learning experience like no other.
The LMS will lessen the strain on the current learning process by employing AI Analytics to enhance the learning experience. The issue of overcrowded classrooms will cease to affect the learning experience because the LMS will offer an individualized approach to teaching that adapts to each learner’s strengths and areas for progress.
Adopting gamification elements to create an enjoyable and productive platform, the LMS will drive access to quality STEM education and education in general, to schools and communities in scale. This LMS is developed not just to enhance students' technical skills but also to prepare them for future employment in an increasingly digital job market.
The organization envisions teachers using the platform’s data analytics to curate teaching strategies that will be individually effective to every learner while leveraging on the insight given into the learners’ progress, engagement, and cognitive responses. The platform has workspaces with virtual robotics kits/equipment creating a more sustainable model for providing scalable
access to robotics training. Applications of the LMS in the education field are numerous since it can be configured to cater to multiple nuanced environments.
Following the LMS is an educational robotic kit developed using an Arduino board as the brain the system (controlling various functions and attachments). Together with the LMS, the kit can be tailored to suit learners of different levels, with simple tasks for younger students and more complex algorithms for advanced learners.
Looking ahead, United Siyafunda NPC aims to expand its reach further into rural areas where access to technology is even more limited. The organization's vision includes not only increasing its impact but also erecting community digital hubs while working to enable access to EdTech resources in under-resourced and marginalized communities and schools.
United Siyafunda NPC stands at the forefront of a movement aimed at bridging the digital divide in South Africa's under-resourced communities. Through effective leadership by Veli Mokhabodi and Dieketseng Malekele, coupled with innovative solutions to overcome challenges like high equipment costs, the organization is making significant strides toward ensuring equitable access to STEM education.
As it continues to empower youth through coding and robotics training, United Siyafunda NPC not only enhances individual lives but also contributes to building a more skilled workforce capable of driving economic growth in South Africa. The journey towards achieving these goals is ongoing, but with each step forward, United Siyafunda NPC reaffirms its commitment to fostering a brighter future for all South Africans through technology and education.
...continued from page 55
The training does not stop there though. Each business has specific terminology that needs to be aligned with to accommodate FAQs, rules, product info and the like.
●In our experience though, we find it advantageous to explicitly leverage product-specific language with our models to really ensure that the interactions are meaningful and the customer experience is as expected. Take the word balance again - for an investment account it has a positive meaning, but the same can’t be said for a home loan or other forms of credit.
These are the tips of many, many icebergs.
This level of sophistication requires not only skill, but a platform robust enough to manage any mode of input and marry them with the correct response. This requires language detection, sentiment analysis, intent mapping, integrations and much more. Doing the same for other languages like Zulu or Afrikaans is therefore not just a simple exercise of translation. Further work needs to be done in order to leverage the initial investment and effort.
The above training of these LLMs needs to cater for text-based interactions as well as factor in alternative modalities –voice, vision or even virtual reality.
It is a big, big undertaking, and you can now see why it’s not as simple as plugging the likes of ChatGPT into your brand platforms.
We have just taken a new product to market called Helm Gen, which makes use of a pivotal advancement in the realm of LLMs called RetrievalAugmented Generation (RAG). Once we have unpacked the problem, investigated and addressed all of the considerations above, Helm Gen uses RAG to provide a Generative AI solution that operates within the bounds of brand content, while limiting ‘hallucinations’.
The reason I mention this now is that the considerations I have raised here are not intended to dissuade anyone from using these GenAI solutions – all we are trying to do is ensure that businesses understand the investment required to truly realise the benefits of Generative AI for business.
Let’s chat.
Chat to Dawood by booking an appointment here
To find out more about Helm, visit their website.
/ By Rob Floyd /
The African Union announced the continental AI strategy on August 9, 2024. The framework was urgently needed as some countries already have AI strategies or policies in place, and others are moving quickly to establish frameworks to guide the governance of, and use of, artificial intelligence.
On the heels of the new strategy, key stakeholders recently reflected on what is needed to ensure the strategy is implemented effectively and the continent avoids a patchwork of strategies, legislation, and regulation. During the summit on “Shaping Africa’s Artificial Intelligence Future: Strategies for Inclusive and Collaborative National Artificial Intelligence Policy Development” on August 20, hosted at the African Union participants sought to accelerate unified approaches to policy development, align investment strategies, ensure inclusive AI, and foster collaboration for AI development.
That implementation of the continental strategy for AI will rely on a number of interrelated foundational elements, from digital skills to digital infrastructure to related continental strategies such as the AU Data Policy Framework and the African Continental Free Trade Area Agreement. But equally important, successful implementation of the AU AI Strategy willrely on collaboration between African governments, industry, academia and policy institutes, regional and sub-regional institutions, citizens and development partners.
Each of these actors has a unique role to play, and collaboration is needed between all of them. If this collaboration does not
happen quickly and lead to well-informed policy, accelerated innovation, the creation of jobs, and improved development outcomes, many countries in Africa will likely fall behind global trends and norms for artificial intelligence.
The AU strategy is clear in calling for collaboration. One of the high level recommendations isto “foster collaboration between the government, private sector, academia, civil society, andregional and international organizations to skill up and build startups’ capacity to deploy inclusive and development-orientated AI solutions and systems in Africa”. It also outlines numerous other recommendations that will require partnership and coordination such asensuring regulatory and legislative harmonization, engaging in international dialogues to share best practices, promoting regional data pools and creating open science platforms.
There is a wide array of partners committed to support implementation of the continental AI strategy to include AU organs such as AUDA-NEPAD and the AU Working Group on AI chaired by Egypt; development partners such as the EU, Germany and the United Kingdom; civil society and citizen platforms such as Omdena; technology firms such as Meta, Microsoft and Google;and research institutions such as Research ICT Africa and the African Center for EconomicTransformation (ACET).
The cross border nature of AI and its inevitable transnational impact renders regional cooperation a necessity rather than a luxury. Africa is projected to be the most affected region in the world by AI. This leaves governments with no other option than to intensify their collaboration, which must also include platforms to gather citizen input and understand concerns related to AI.
Given that only a few African countries have AI strategies or policies in place, there is a real opportunity for peer learning among nations, building upon lessons learned, and sharing experiences. In some ways, industry has the most to off er in this regard given the global technology fi rms have a wealth of experience from around the world. They also bring different perspectives. For example, some firms will have more expertise in infrastructure and interoperability, while others will have helped craft policies on data protection or connectivity.That said, it is critical that the local private sector is also crowded into the dialogue as they will reflect the contexts and challenges in each country. As well, innovators and entrepreneurs will likely have the best understanding of local demand. The continental AI strategy places a strong focus on how AI start-ups can be incentivized and how the investment environment can bestrengthened in specific markets.
At the same time, governments have an obligation to engage citizens. While there are often narratives about the role of civil society and citizens in policy making, in reality they are not always included sufficiently in the policy making process. This is particularly the case when itcomes to rural communities and disadvantaged groups. In some cases, umbrella groups canbe engaged, and in other instances local NGOs involved in digital skills can contribute. There are also citizen platforms such as Omdena that bring together data scientists and experts to craft AI solutions to development challenges. Across the continent there is a vibrant network of think tanks and policy institutes supporting AI development. These can be further leveraged to help inform AI policy. Some such as NIYEL in Senegal are focused on strengthen responsible AI in a particular region such as West Africa; others such as the African Technology Policy Studies Network (ATPS) is focused on sectors such as agriculture; while some such as the African Center for Economic Transformation (ACET) support development of AI solutions to inform economic policy. These, and other, think tank scan provide independent analysis, support peer learning, and advocate for AI policy that leadsto the greatest outcomes.
The authors, writing in their personal capacities as professionals involved in AI development,argue for action in six key areas:
1. Rapid opportunities for peer learning among governments in safe spaces where both successes and challenges can be addressed.The AI agenda is moving quickly, therefore governments need strategies and policies in place, but also run the risk of developing legislation that is not fully informed. This should include the establishment of an African/AU AI observatory to encourage sharing of best practices and experiences. This will serve as a guideline/benchmark for countries who are yet to develop the AI national strategy.
2. A network of centers of excellence, innovation hubs, AI institutes, and think tanks.To support the observatory and peer learning opportunities, priority should be given toestablishing these centers and groups. The African Union Commission, the AU working groupon AI and AUDA-NEPAD have particularly roles to play in supporting peer learning but they should also leverage African think tanks and development partners.
3. Greater citizen involvement at the national and regional level to inform AI policy. In particular, citizens can help cocreate solutions, ensure that AI policies are inclusive, andinform ethical and responsible artificial intelligence policy.
4. Collaborative input between government and Industry—both global and local. Industry should provide input to the policymaking process that builds upon international experience and local needs and demand from citizens and business. Innovators and entrepreneurs can help policymakers understand. Governments need to be proactive in calling upon the privatesector to solicit inputs.
5. A permanent review body. A review committee comprised of AU member states representing sub-regions, AU specialized agencies and regional economic communities tomonitor implementation of the AU Artificial Intelligence Strategy should be established. The review committee should report out to an AU Commission African Multi stakeholder AI Assembly.
6. Continent-wide capacity building programs. Such programs should be established to cover key aspects of AI policy, including in-country campaigns to raise awareness.
The time for action is now. More than seventy countries (and regions such as the EU) have developed AI strategies, policies and frameworks. Many of these strategies continue to evolve or are in early stages. But to date only a few African countries have strategies in place, and even fewer have parliament-approved policies.
One of the biggest challenges to AI adoption in most of African economies is changing culture.Many citizens are cautious and skeptical towards embracing AI. ACET’ research has also shown there is often a lack of digital leadership within government. The political leadership in Africashould champion continental partnership frameworks to address the vital obstacles to applying AI systems. The African Working Group on AI established in 2019 with the mandate to develop and implement an African continental AI strategy demonstrates one of the successful models such cooperation.
As governments across the continent continue to pursue economic transformation, AI can be akey driver of such transformation. There are estimates that AI is worth nearly $20 trillion to theglobal economy and $1.5 trillion to Africa by 2030, while productivity gains may equal morethan 15% to the global GDP. As with any innovation or new technology, there will be losers in the economy, and there is the potential for bad actors – but the benefits will certainly outweigh the costs over the medium term. That said, African governments must quickly develop harmonized strategies and approaches that build upon the continental AI strategy if they are to capitalize on those gains.
We already see tremendous use of AI in Africa, from tools to address climate change tosolutions for healthcare and education. It will, in the coming years, become part of the economic policymaking process and become fully integrated into our daily lives. Effective collaboration to develop AI, and the policies to regulate it, are critical to fully leveraging its benefits. That collaboration has begun, but must be accelerated.
Rob Floyd: Director for Innovation and Digital Policy, African Center for Economic Transformation
Amr Farouk Safwat: Manager of African Organizations, Ministry of Communications and Information Technology, Egypt
Souhila Amazouz: Senior Tech Policy Officer, African Union Commission
Mlindi Mashologu: Deputy DirectorGeneral, ICT Information Society and Capacity Development Africa Development at Department of Communications and Digital Technologies, South Africa
James Elieta: Vice President for Partnerships, Omdena
Joseph Ishie: Client Advisory & Business Development Manager, Nina Jojer
/
Deputy Defence Minister Bantu Holomisa delivering the keynote address at the Symposium on Responsible AI in the Military Domain.
The Global Commission on Responsible Use of AI in the Military domain (GC REAIM) has successfully concluded the South African leg of its world tour in Stellenbosch (11-14 Nov 2024) co-hosted by the Defence AI Research Unit (DAIRU) located at the Faculty of Military Science of Stellenbosch University.
GC REAIM is an initiative of the Government of the Netherlands that was launched during the 2023 REAIM Summit on Responsible Artificial Intelligence in the Military Domain in The Hague.
It appointed the Hague Centre for Strategic Studies (HCSS) as the Global Commission for this initiative. DAIRU is a centre situated in Saldanha Bay and was launched in early 2024 as the 4th national AI hub by the Department of Communications and Digital Technologies.
The Hon Deputy Minister of Defence and Military Veterans of the Republic of South Africa, Maj Gen (ret) B.H. Holomisa gave the ministerial address at the Symposium Day on 14 Nov 2024. Amongst the distinguished VIP guests were Her Excellency Ms. Helene Rekkers, Consul General of the Kingdom of the Netherlands in Cape Town, and Professor M.S. Tshehla, Dean of the Faculty of Military Science of Stellenbosch University located at the South African Military Academy.
Invited guests from the South African National Defence Force, academia, government, industry, and civil society heard keynotes and panel discussions on three important topics of relevance:
The keynote address was given by Mr Zane Cleophas, followed by a panel discussion
moderated by Dr Tim Sweijs. This first session explained what AI is, what its use cases are in the South African National Defence Force, and the legal, moral and ethical implications of its use. Challenging us to consider the harm caused by armed conflict and who benefits from the investments made in AI.
The keynote address was given by Col (ret) Dr Piet Bester, followed by a panel discussion moderated by Ms. Noelle van der WaagCowling. This session unpacked the legal, moral and ethical considerations of the use of AI in the military, specifically the use of Lethal Autonomous Weapons Systems and ensuring that responsible design of AI is done cognizant of International Law as it relates to armed conflict.
The final keynote address was given by Dr Moses Khanyile, followed by a panel discussion moderated by Ms Sofia Romansky. This session emphasized compliance with international law and the geopolitical dimensions of the design and use of AI, particularly how well state actors and industry were doing in terms of compliance.
/ Read original article here /
The global conversation surrounding the ethical integration of Artificial Intelligence (AI) in military operations took a significant step forward at the Symposium on Responsible AI in the Military Domain.
This was held in Stellenbosch, South Africa, on 14 November. The event, co-hosted by the Global Commission on Responsible AI in the Military Domain (GC REAIM) and the Defence AI Research Unit (DAIRU) of Stellenbosch University, gathered military practitioners, academics, policymakers and civil society leaders to exchange interdisciplinary perspectives on AI’s role in defence.
The Symposium was an extension of the GC REAIM conference, which had run for three days prior, delving into the intersection of AI and defence systems. This particular day focused on the ethical challenges and opportunities AI presents within the military context, addressing everything from technological foundations to the implications for peace, security and governance.
The initiative began with the first summit in February 2023 as the Commission aims to raise awareness, foster understanding and deliver a comprehensive guidance report on responsible military AI. Key issues under discussion include addressing rapid technological advancements, ensuring AI use aligns with peace and security goals, embedding ethics in decision-making and developing regulatory frameworks grounded in international law.
The highlight of the event was the keynote address by South African Deputy Minister of Defence, Major General (Retired) Bantu Holomisa. Holomisa discussed the transformative potential of AI in enhancing military capabilities, with particular emphasis on its application in border monitoring and counteracting illicit activities. He underscored that while AI presents substantial opportunities for improving security, efficiency and decisionmaking, it also brings forward ethical responsibilities.
“Artificial intelligence is not merely a tool for military advancement; it is an opportunity to reinforce our commitments to peace and stability,” Holomisa stated.
His remarks were rooted in South Africa’s broader strategy for responsible AI deployment, which includes the National Artificial Intelligence Policy Framework and the establishment of DAIRU, designed to promote local research and ensure the ethical use of AI in defence systems.
Holomisa also highlighted the challenges posed by low-cost, off-theshelf technologies, which can undermine advanced defence systems. He emphasised the need for adaptive procurement strategies and robust oversight mechanisms to protect national security interests.
Referencing British politician Tom Tugendhat, Holomisa stated: “The economies of war have changed and lowcost technology is now able to undermine advanced equipment. Defence leaders therefore must adapt swiftly, enhancing oversight and regulation of artificial intelligence to ensure national security keeps pace with technological shifts.”
“Reforming procurement systems is essential to address these emerging challenges,” he added.
He emphasised that AI should be used to bridge digital divides and foster equity, stressing the importance of international collaboration for successful AI integration in the military.
The symposium featured several discussions on the ethical implications of integrating AI into defence systems. AI’s ability to enhance decision-making, predictive analytics and autonomous operations in military contexts bring numerous advantages. However, experts raised critical concerns about the misuse of autonomous weapons and the ethical challenges of entrusting life-or-death decisions to machines.
A panel of experts explored these concerns, discussing the need for human oversight in autonomous systems, especially those involved in lethal actions. The consensus was clear: AI should assist, not replace, human decision-making in situations where moral judgments are essential. Furthermore, there was a call for the development of international governance frameworks to regulate AI use in military settings, ensuring that its deployment aligns with humanitarian and legal principles.
One of the central ethical concerns discussed was the potential use of lethal autonomous weapons, which has sparked global debate. While AI can significantly improve targeting precision and reduce collateral damage, its deployment in combat raises moral questions about accountability, transparency and the risks of unforeseen consequences. The need for clear guidelines and regulatory frameworks to govern AI’s use in warfare was emphasized by several speakers who underscored the importance of ethical frameworks in mitigating such risks.
The issue of governance was also a major theme at the Symposium. The dual-use nature of AI (where the same technology can be applied to both civilian and military domains) complicates efforts to regulate it effectively. Experts discussed the need for international cooperation to create comprehensive governance structures that balance innovation with the responsible use of AI in defence.
Proposed solutions included the establishment of global AI conventions, similar to frameworks for nuclear or chemical weapons, to set clear guidelines and protocols. Additionally, the role of accountability across the entire AI value chain, from developers to end-users, was seen as critical to ensuring AI systems are used ethically and in accordance with international law.
However, the potential consequences of a State or non-State actor’s deviation from ethical AI development standards are yet to be fully understood.
The Symposium served as a crucial platform for shaping the future of AI in defence, where discussions on ethical considerations, governance and international collaboration will continue to drive the global dialogue on AI’s role in military applications.
Future plans include institutionalising these efforts, presenting a draft Strategic Guidance Report during the NATO summit in June 2025. The Strategic Guidance Report will contain short- and long-term recommendations on norm development on responsible design and use of Al in the military domain. The ultimate goal is to submit the report to the UN General Assembly, advocating for actionable frameworks to ensure responsible AI integration in military contexts.
Africa's only AI Trade & Innovation Magazine
BOLSTERING AI FOR GOOD UN ITU Innovation
Across Africa
TRANSLATE adds
25+ African languages
SAAIA Raises
Concerns about AI Model Training
SUPERMICRO
Preparing for an AI driven world
ARTIFICIAL INTELLIGENCE & 4IR COMMUNITY WITH SYNAPSE MAGAZINE
Published Quarterly
O cial Publication of AI Expo Africa
RATE CARD
Company Listing (logo, company description & hyperlink) R2500
1/4 Page (includes 1/4 page editorial) R3500
1/2 Page (includes 1/2 page editorial) R4500
Full Page (includes full-page editorial feature) R6500
Double Page Spread R10,000
* All rates exclude VAT & agency commission. Rates are based on casual advertising. Discounted rates are available for longer ad / editorial runs
Synapse Magazine is Africa’s first and only business quarterly publication covering developments across the continent in Artificial Intelligence (AI), Data Science, Robotic Process Automation (RPA) and Fourth Industrial Revolution (4IR) smart technologies.
Synapse offers industry executives, practitioners, investors and researchers relevant news, in-depth analysis, and thought leadership articles on trends around 4IR innovation and digital transformation in industries that include banking, retail, manufacturing, healthcare, mining, agriculture, education, and government, among others.
With its insights, interviews and case studies, the magazine aims to be a voice for African 4IR practitioners, researchers, innovators, thought leaders, and the wider African AI community.
Since its launch in 2018, Synapse has amassed a combined readership of 31,300 across the Issuu platform (on which it is published), the AI Media Group’s email database, the AI Expo Africa Community Group on LinkedIn and the AI Media Group’s social media channels where the magazine is distributed. It also links to AI TV, Africa’s only dedicated YouTube streaming channel focused on 4IR business users and trade.
Advertising and artwork to be supplied as a high resolution PRESS-ready PDF of at least 300dpi. Art and editorial features to be submitted to: daniel.mpala@aiexpoafrica.com AD SIZE
Over the years the magazine has established a significant following across Africa as well as globally, with readers from as far afield as the North America, South America, Europe and Asia. This makes Synapse a great marketing platform for startups and established tech companies to reach a broader community of buyers, investors and partners.
Readers around the world
