The Future Engineer Is Not Replaced by AI. The Future Engineer Is the One Who Knows How to Use It.
How Mctaba Labs Is Rethinking Software Engineering Education for the AI Era, and Why Africa Can't Afford to Wait
Today, I sent an internal memo to every student and staff member at Mctaba Labs. The subject line was blunt: AI Is Now Part of the Curriculum From Day One.
Not month three. Not as an elective. Not as a "nice-to-have" module tacked on at the end of a bootcamp. From Day One. Starting April 13th.
Some people will read that and think it's reckless. That we're moving too fast. That students need to "learn the basics first" before touching AI tools.
I disagree. And I'm not alone.
The World Is Already Moving
Jack Dorsey recently published a piece titled From Hierarchy to Intelligence, laying out how Block is fundamentally restructuring its entire company around AI. Not as a productivity hack. Not as a chatbot bolted onto customer support. As the actual coordination layer of the organisation.
His argument is striking: for two thousand years, from the Roman Army to modern corporations, we've organised humans into hierarchies because that was the only way to route information at scale. A leader manages three to eight people. Those people manage three to eight more. Information flows up. Decisions flow down. That's the model. Every company on earth runs on it.
Dorsey's bet is that AI breaks that constraint entirely. At Block, they're building what he calls a "company world model", a system that maintains a continuously updated picture of the entire business, replacing the information-routing function that middle management has performed since the Prussian military reforms of 1806. Individual contributors get context directly from the model. Cross-functional problems get owned by Directly Responsible Individuals. Managers become player-coaches who build things and develop people, not route information.
This isn't theoretical. Block is doing it now.
Meanwhile, Andrej Karpathy, former head of AI at Tesla, co-founder of OpenAI, recently shared how he builds personal knowledge systems using nothing more than three folders and a schema file. No fancy apps. No complex databases. Just raw material, an AI-maintained wiki, and a simple set of rules the AI follows to organise everything. The system compounds: every question you ask makes the next answer better. Every source you add deepens the model's understanding.
The lesson from both of these leaders is the same: AI is not a feature. It's an operating system for how work gets done.
What This Means for Africa
Here's what keeps me up at night.
Africa has the youngest population on the planet. The median age in Kenya is 20. In Nigeria, it's 18. We have more people about to enter the workforce than at any continent in history. And the skills they need to thrive are changing faster than most institutions can update a syllabus.
Most coding bootcamps and university programmes across the continent are still teaching the 2019 version of software engineering. Write code. Pass tests. Get a job writing more code. Maybe learn a framework. Maybe deploy something.
That model is already obsolete.
Companies are no longer hiring developers just to write code. They're hiring engineers who can design systems, ship products, maintain infrastructure, optimise costs, think about security, and work dramatically faster using AI tools. The most valuable engineer in 2026 is not the one who types the fastest. It's the one who decides what should be built, how it should be built, and how it should scale.
That's what Mctaba Labs is training people to become.
Our Philosophy: AI-Integrated, Not AI-Dependent
When I wrote the memo, I was deliberate about one thing: we are not replacing fundamentals.
Students will still learn the basics of coding, algorithms, data structures, system design, databases, security, testing, and production engineering. That doesn't change. What changes is that AI becomes a tool used across all of those disciplines from the start, the same way a carpenter doesn't learn to use a hammer in month six.
Our first two weeks now cover AI development tools setup, prompt engineering fundamentals, AI ethics, and how to use AI within a developer workflow. Weeks three and four go deeper: how models actually work (practically, not academically), where AI is strong and where it fails, how to verify and test AI output, and critically, when not to use it.
From month two onwards, AI is integrated into everything: labs, projects, testing, documentation, debugging, and deployment workflows.
But here's the part that matters most: for every aspect of AI usage in projects, students must write an accompanying markdown file explaining how they used AI. This isn't busywork. This is the skill. Understanding what the tool did, why you used it, and whether the output is correct, that's the difference between an engineer and someone who copies and pastes from a chatbot.
We took a page from the Karpathy playbook here. His approach to knowledge management is built on a simple principle: the AI organises, but the human directs and validates. The schema file, the rules that tell the AI how to behave, is written by the human. The judgment is human. The speed is AI. That's the balance we're teaching.
The Risks Are Real. We're Not Ignoring Them.
I laid out four risks in the memo, and I want to be transparent about them here too.
Over-reliance on AI is the biggest one. If students learn to prompt before they learn to think, we've failed. That's why we have manual coding assignments, code reviews, debugging challenges, and system design assessments that AI can't do for you.
AI giving wrong answers is a certainty, not a possibility. We train students to test, validate, and question outputs, not blindly trust them. This is perhaps the most important engineering skill of the next decade.
Security and privacy are non-negotiable. Every project includes security modules and safe data practices.
Skipping fundamentals is the trap everyone worries about. Our assessments test understanding, not copy-pasted solutions.
We're not removing difficulty. We're shifting the difficulty from typing code to thinking like an engineer.
Why This Is a Business Opportunity, Not Just an Education Play
Let me be direct about the commercial reality.
Jack Dorsey's essay ends with a stark warning: if a company's answer to "what does your organisation understand that is genuinely hard to understand?" is nothing, then AI is just a cost-cutting story. You trim headcount, improve margins for a few quarters, and eventually get absorbed by something smarter. But if the answer is deep, if you're building compounding understanding, AI reveals what your company actually is.
At Mctaba Labs, our compounding advantage is this: we are building the first generation of African engineers who think AI-natively. Not engineers who learned AI as an afterthought. Engineers who have never known a workflow without it. Engineers who can build AI-powered products, design agentic systems, deploy automated workflows, and ship production-grade software, all within the context of African markets, African payment infrastructure, and African business problems.
That's not something you can replicate by watching a YouTube tutorial. That's an ecosystem. And ecosystems compound.
Every student who graduates from Mctaba and builds a product is proof of concept. Every product they ship is a case study. Every company that hires them validates the model. Every founder who emerges from this programme expands the network. This is the flywheel.
What Graduation Looks Like Now
Our capstone projects have changed. By the time you finish at Mctaba Labs, you should be able to:
Build and deploy full production systems
Use AI tools as a seamless part of your workflow
Design system architectures from scratch
Review and debug AI-generated code with confidence
Integrate AI into applications where it adds genuine value
Think about scalability, cost, and security as first-order concerns
Ship real products, not assignments
Capstone projects now include AI-powered features, automated workflows and agents, production deployments with monitoring and logging, and real-world architecture decisions.
We are not producing people who can pass coding tests. We are producing engineers who can build companies.
Many Will Wait. We Won't.
Many schools will wait for AI to stabilise before adding it to their programmes. Many curricula will remain unchanged for years. Many programmes will debate internally while the industry moves on without them.
We are not going to wait.
Technology does not wait. Industry does not wait. Opportunities do not wait.
Karpathy keeps his knowledge system "super simple and flat" because the tool doesn't matter; the thinking does. Dorsey is rebuilding a $40 billion company around AI because the old structures can't keep up. Both of them are operating on the same insight: the organisations and individuals who integrate AI into how they think and work, not just what tools they use, will define the next decade.
Africa has a window. The youngest workforce. The fastest-growing tech ecosystems. The most urgent problems to solve. And now, for the first time, the tools to solve them at a speed and scale that was previously impossible.
Mctaba Labs is walking through that window.
If you're ready to walk through it with us, we're here.
Bonaventure Ogeto is the founder of Mctaba Labs, an Africa-First technology company based in Nairobi, Kenya, building across AI, automations, software engineering, and payment infrastructure. Mctaba's Software and AI Engineering Career Launch Programme is currently enrolling.