Resource Library

Go deeper.

Case studies from every module with sources. Books and reports that sharpen your thinking. Podcasts and channels worth following. Curated for professionals building real AI judgment.

Reading List

Books and reports for learners who want to go beyond the course. Each one earns its place.

Books
AI Snake Oil cover
Book · 2024
AI Snake Oil
Arvind Narayanan & Sayash Kapoor
The clearest explanation of what AI can and can't do — and why most AI claims are overblown. Essential for anyone navigating AI hype at work.
→ aisnakeoil.com
Weapons of Math Destruction cover
Book · 2016
Weapons of Math Destruction
Cathy O'Neil
How algorithms harm the poor, reinforce discrimination, and threaten democracy. The classic case for why AI judgment matters.
→ Find on Takealot or your library
Race After Technology cover
Book · 2019
Race After Technology
Ruha Benjamin
How technologies designed by some, for some, become tools of racial inequity. Essential context for understanding why AI in Africa isn't a neutral import.
→ ruhabenjamin.com
Power and Prediction cover
Book · 2022
Power and Prediction
Agrawal, Gans & Goldfarb
The business economics of AI: why AI is about shifting who makes predictions, what that does to power, and how professionals need to adapt.
→ Find on Takealot or your library
The Alignment Problem cover
Book · 2020
The Alignment Problem
Brian Christian
What it takes to build AI that does what we want — and the many ways it goes wrong. Accessible deep dive into AI failure modes, for the curious non-engineer.
→ brianchristian.org
You Look Like a Thing and I Love You cover
Book · 2019
You Look Like a Thing and I Love You
Janelle Shane
Funny, zero-pressure intro to AI. Explains why AI can do impressive things while also doing deeply stupid, bizarre things. Helps you lose the "AI is magic" feeling fast — and laugh while doing it. Start here if you're completely new.
→ janelleshane.com/book
Artificial Intelligence: A Guide for Thinking Humans cover
Book · 2019
AI: A Guide for Thinking Humans
Melanie Mitchell
The best next step once you want a more serious but still readable explanation. Walks through neural networks, computer vision, language, and the gap between current AI and human intelligence — no math or coding required.
→ Find on Takealot or your library
How AI Thinks cover
Book · 2024
How AI Thinks
Nigel Toon
More modern and ChatGPT-era. Explains the basic story of how machine learning developed into generative AI, what these systems are doing, and why people worry about control and power. The bridge from "I understand the basics" to "I understand the current moment."
→ Find on Takealot or your library
AI Needs You cover
Book · 2024
AI Needs You
Verity Harding
Less about the technical machinery and more about the human side: who gets to decide how AI is used, what rules matter, and why normal people — not only computer scientists — should be in the conversation. A good balance against reading only Silicon Valley "AI is inevitable" perspectives.
→ Find on Takealot or your library
Reports & Policy Documents
FSCA AI in Financial Sector report
Report · November 2025
AI in the South African Financial Sector
FSCA & Prudential Authority
The first comprehensive look at how SA's major financial institutions are deploying AI and where the gaps are. What your regulator already knows about AI in your sector.
→ Read the summary
ICASA State of ICT Sector report
Report · 2025
State of the ICT Sector in South Africa
ICASA
Annual review of SA's telecoms and digital landscape. Essential context for how AI is entering the sector most actively adopting it — and what the regulator is watching.
→ Download PDF (ICASA)
AU Continental AI Strategy document
Policy Document
Continental AI Strategy for Africa
African Union
The continental framework for AI development in Africa. Shows how SA's policy fits into the broader African picture and where Africa's governance priorities differ from the US or EU.
→ African Union →
UNESCO AI Ethics Recommendation
Policy Document · 2021
Recommendation on the Ethics of AI
UNESCO
The global consensus document on responsible AI — adopted by 193 countries. The foundation for understanding what "ethical AI" actually requires in practice.
→ UNESCO.org →
CIPESA State of Internet Freedom report
Report · 2025
State of Internet Freedom in Africa
CIPESA
Covers AI-enabled surveillance, deepfake disinformation, and digital rights across 25 African countries. The most comprehensive annual review of AI vs. freedom on the continent.
→ cipesa.org →

Podcasts

Worth adding to your commute. All of these are genuinely good — no filler.

Hard Fork podcast cover
Weekly · News & Culture
Hard Fork
New York Times — Kevin Roose & Casey Newton
The most accessible weekly podcast on AI and tech. Conversational, current, and honest about both the hype and the stakes. Good for staying current without getting lost in jargon.
→ nytimes.com/column/hard-fork
Eye on AI podcast cover
Weekly · Research & Industry
Eye on AI
Craig Smith
Weekly interviews with AI researchers, ethicists, and practitioners. Goes deep without losing the plot. Good for building real understanding of where AI is — not just where people say it's going.
→ eye-on.ai
Practical AI podcast cover
Weekly · Applied & Professional
Practical AI
Changelog — Daniel Whitenack & Chris Benson
For professionals who want to apply AI at work, not just understand it theoretically. Business-focused, non-technical, grounded in what actually gets implemented — not just what gets announced.
→ changelog.com/practicalai
80,000 Hours Podcast cover
Long-form · AI Safety & Ethics
80,000 Hours Podcast
80,000 Hours
Long-form conversations on AI safety, governance, and the long-term consequences of AI development. Serious and evidence-based. Dense, but the episodes on AI risk and AI policy are worth the time.
→ 80000hours.org/podcast
Tech Cabal podcast cover
Weekly · African Tech
The TC Daily
Tech Cabal
The African tech landscape from an African perspective. Covers AI, digital finance, and startup ecosystems across the continent. Essential for staying current on how AI is actually playing out in Africa.
→ techcabal.com/podcast
The AI Daily Brief
Daily · Podcast + Newsletter + Website
The AI Daily Brief
Nathaniel Whittemore
Daily AI news and analysis — as a podcast, a newsletter, and a website. One of the best ways to stay current on what's actually happening in AI without drowning in noise. Subscribe to the newsletter if you can only pick one.
→ theaidailybrief.com
Data Skeptic podcast cover
Podcast · Statistics & AI
Data Skeptic
Kyle Polich
The best "I want to understand this without being treated like I already have a PhD" option. Covers statistics, machine learning, data science, and AI with a scientific-skeptical lens instead of pure hype. Good for learners who want the rigour without the jargon.
→ dataskeptic.com
The Cognitive Revolution podcast cover
Podcast · AI Strategy & Safety
The Cognitive Revolution
Nathan Labenz
Good for understanding what AI systems can realistically do, where companies are actually using them, and what the big arguments are around agents, reasoning, work, and safety. More conversational than instructional — closer to a front-row seat at the serious AI conversation.
→ cognitiverevolution.ai

YouTube Channels

Seven channels worth your time. Start with one video from each and go from there.

3Blue1Brown channel icon
YouTube · Visual Explainers
3Blue1Brown
Grant Sanderson
The best visual explanation of how AI and machine learning actually work. His "Neural Networks" series is the clearest explanation of what's inside the black box — no coding required, just curiosity. Start here.
→ youtube.com/@3blue1brown
Two Minute Papers channel icon
YouTube · AI Research
Two Minute Papers
Károly Zsolnai-Féhér
Short, excited summaries of the latest AI research papers. Two minutes to understand what just changed in the field. Good for staying current on AI capabilities without reading academic papers yourself.
→ youtube.com/@TwoMinutePapers
Fireship channel icon
YouTube · Fast Takes
Fireship
Jeff Delaney
Fast, entertaining explainers on AI tools and tech — usually 100 seconds to 10 minutes. If something is happening in AI right now, Fireship has probably made a video about it. Great for keeping up without going deep.
→ youtube.com/@Fireship
Andrej Karpathy channel icon
YouTube · Deep Technical
Andrej Karpathy
Former Tesla AI Director & OpenAI researcher
The most respected technical educator in AI. His "Let's build GPT from scratch" video is the clearest explanation of how large language models work. Rewards patience — this is the one to watch when you want to truly understand what's inside the tools you use every day.
→ youtube.com/@AndrejKarpathy
Yannic Kilcher channel icon
YouTube · AI Research Papers
Yannic Kilcher
Yannic Kilcher
Deep walkthrough of landmark AI research papers — including the original Attention Is All You Need paper and GPT-series models. More technical than Two Minute Papers but far more thorough. Good for learners who want to understand the actual science.
→ youtube.com/@YannicKilcher
StatQuest with Josh Starmer channel icon
YouTube · Statistics & ML Foundations
StatQuest with Josh Starmer
Josh Starmer
The best resource for the pieces underneath AI: probability, statistics, loss functions, gradient descent, neural nets, decision trees, transformers. His videos are slow, clear, and genuinely patient. Good for learners who want to understand what's actually happening when AI learns.
→ youtube.com/@statquest
DeepLearning.AI channel icon
YouTube · Structured AI Education
DeepLearning.AI
Andrew Ng
More course-like than entertainment. The most structured serious AI education available free on YouTube. Good once you want to move from understanding concepts to learning how machine learning and modern generative AI systems are actually built and used.
→ youtube.com/@deeplearningai

Course Case Studies

Every real-world AI case referenced across the modules, with links to the original sources. Use these to revisit a case after a session or share with a colleague.

Module 1 Government · South Africa · April 2026
SA Government AI Policy Hallucination Scandal
South Africa's Department of Communications published an 86-page national AI policy draft containing fabricated academic citations — journals that don't exist, papers that were never published. The policy was withdrawn within weeks. The people responsible for governing AI in SA didn't have the skills to verify AI output before releasing it publicly.
Read: The Conversation →
Module 2 Finance · SA / Nigeria / Kenya · 2025
Credit Scoring: The 37% Gender Penalty
An audit of 10 credit scoring algorithms across South Africa, Nigeria, and Kenya found that women-led SMEs faced a 37% underfunding penalty compared to similarly-positioned male-led businesses — through sector categorization bias and language pattern scoring in the models. No individual loan officer decided to discriminate. The algorithm did it invisibly.
Read: Advanced Research Journal →
Module 2 Finance · Kenya · Ongoing
Kenya Mobile Lending: When Access Becomes the Score
Kenya's digital lending apps score creditworthiness from phone usage data, browsing history, and social media activity. People with less digital footprint get worse credit scores — not because they're less creditworthy, but because they have less access to data, electricity, and smartphones. The AI is measuring infrastructure inequality and calling it credit risk.
Read: PMC / Frontiers →
Module 3 Law Enforcement · South Africa · Ongoing
SA Predictive Policing: Apartheid Data, Modern Algorithms
Predictive policing tools in South Africa were trained on historical crime data reflecting decades of apartheid-era over-policing of Black townships. The AI learned to flag the same communities. More policing followed. More arrests happened. The model was "confirmed." A feedback loop that encodes structural racism into an algorithm without anyone programming discrimination into it.
Read: PMC / Frontiers →
Module 3 Biometrics · Multi-country · 2018–present
Facial Recognition: Built for Some, Deployed for All
A landmark study tested commercial facial recognition APIs used across Africa in banking, border control, and law enforcement. Dark-skinned women were misclassified at substantially higher rates than light-skinned men. These systems were validated for Western populations and deployed in African contexts without re-validation. They are still in use.
Read: PMC / Frontiers →
Module 4 Media & Elections · South Africa · May 2024
SA 2024 Election Deepfake Videos
During South Africa's May 2024 elections, AI-generated deepfake videos circulated across social media showing political figures making statements they never made. Most ordinary voters had no way to distinguish real from synthetic footage. South Africa had no deepfake regulation at the time, and most media organizations lacked verification protocols.
Read: CIPESA State of Internet Freedom in Africa →
Module 5 HR & Legal · South Africa · 2025
AI Recruitment and the Employment Equity Act
SA law firm Cliffe Dekker Hofmeyr issued a 2025 legal alert: if AI recruitment tools produce discriminatory outcomes, SA employers are liable under the Employment Equity Act — even if a human never made the decision. "The algorithm decided" is not a legal defence in South Africa. The pre-litigation window is open now.
Read: Cliffe Dekker Hofmeyr Employment Law Alert →
Module 5 Finance · South Africa · Ongoing
MyBucks: When "It Works" Isn't Enough
A South African fintech's AI credit scoring reduced loan defaults by 18% — a measurable business success. But the same model used opaque decision-making that couldn't be audited for individual fairness. Some customers were denied credit they deserved, with no explanation available. Aggregate improvement masked individual harm.
Read: PMC / Frontiers →
Module 6 Healthcare · South Africa · 2024–2025
SA Healthcare AI: Ahead of Validation
South Africa reports above-global-average AI adoption in clinical settings — but the adoption is ahead of the validation infrastructure. Models are deployed before being tested on SA patient populations, disease burdens, and clinical contexts. When AI trained on Western patient data is applied to SA patients, error rates increase for patients who don't match the training profile.
Read: South African Medical Journal →
Module 6 Education · South Africa · Ongoing
Dr Math: What Responsible AI Deployment Looks Like
The CSIR's Dr Math program uses AI-powered maths tutoring to reach South African students in schools without qualified teachers. It works because the purpose is clear (expanding access, not replacing teachers), the context is understood, and the system augments human judgment rather than replacing it. A benchmark for responsible AI deployment.
Read: UNESCO Good Practice →
Supplemental: African AI Success Stories
Supplemental Finance · SA / Kenya / Ghana / Tanzania / Zambia
Jumo: AI Lending to the Unbanked
Founded in Cape Town, Jumo uses AI to extend credit to people with no formal credit history — scoring creditworthiness from mobile usage patterns in partnership with MTN and Airtel. Operating across five African markets, Jumo is Africa-built AI solving a structural financial exclusion problem that traditional banking couldn't reach.
Visit: jumo.world →
Supplemental Civic Tech & Health · Johannesburg-based · Multi-country
Praekelt.org: Mobile-First AI for Public Health
This Johannesburg-founded nonprofit builds AI-enhanced platforms for government health programs across Africa. Their COVID-19 vaccine registration platform served millions of South Africans in 2021. Their WhatsApp-based health communications reach populations where other digital channels don't work. Africa-built, Africa-scaled, Africa-centered.
Visit: praekelt.org →
Supplemental Education · Cape Town-based · SA-wide
Siyavula: SA-Built Adaptive Learning AI
Built in Cape Town specifically for South Africa's CAPS curriculum, Siyavula's adaptive maths and science platform personalizes practice for SA high school students. The Department of Basic Education and several provinces have formal partnerships giving public school students free access. Locally-built AI designed from the ground up for a specific local context — not an import.
Visit: siyavula.com →