Africa doesn't need to build the world's most powerful AI model to benefit enormously from the technology. But it does need the infrastructure, institutions and expertise to make sure those benefits aren't limited to importing services from elsewhere.

The countries building the most advanced AI systems will not be the only ones living with the consequences. A worker in Lagos or Nairobi could lose opportunities because of decisions made in Silicon Valley without ever having a voice in those decisions.

There's something wrong with treating AI governance as a conversation between the US, China and a few technology companies. The economic consequences will reach countries that have little influence over the systems being developed.

If AI can eventually manage companies better than their founders, the question of human relevance won't stop at entry-level jobs. The people who currently own and direct businesses may discover that economic authority and superior decision-making no longer belong together.

We tend to think losing a job is mainly about losing money. For some people, it also means losing status, community, routine and a sense that their contribution matters. Any serious plan for widespread automation has to deal with that human reality.

We keep talking about AI making everything cheaper, but cheaper production doesn't automatically mean cheaper living. A company can cut its costs dramatically without passing those savings to customers. Abundance depends on more than how much technology can produce.

Human agency is easy to overlook when discussing AI because it's harder to quantify than income or productivity. But a person who receives everything they need while having no say over the decisions shaping their future isn't experiencing complete freedom.

YOU ARE NEVER TOO YOUNG TO START! I spent a couple of hours hanging out with my incredible son, Ezekiel, who, at just 16, has built over seven applications. To every young person, Start now. Start with the free resources available to you. But whatever you do, START!

I find it difficult to accept that the only choices are racing to superintelligence or rejecting AI altogether. We can develop useful systems, study their risks, improve safeguards and decide which capabilities deserve greater caution.

One thing I've learnt about digital transformation is that the best technology solves problems people already experience. DocTracker is a good example. It brings project reporting, site activities and progress monitoring into a more structured digital process. #SolutionIsHere

The promise of AI in medicine is genuinely exciting. Imagine fewer families losing loved ones to diseases we currently struggle to treat. It's a future worth working towards, and a good reason to care deeply about getting this technology right.

People keep saying, “AI won't replace you. Someone using AI will.” But that person will eventually be replaced too. Learning to use AI is important, but we should stop pretending like it's a lifetime guarantee of job security. IT'S NOT.

If AI can eventually do the work of ten people, we have two possibilities: ten people get more freedom, or one company gets the productivity of ten people while employing fewer. Technology alone won't decide which future we get.

The more capable AI becomes, the less impressive it is to say you used AI to make something. People care about what you made, whether it's good, and whether it solves a problem. The tool is not the achievement.

You can build a fantastic product and still fail because nobody knows where to find it. Before building your startup, ask yourself: where do my customers already spend their time, and how do I reach them there? Distribution is simply not something you figure out after launching.

People deserve to know what's being built, how far work has gone, and what's happening on site. That's the thinking behind DocTracker, giving Anambra residents better visibility into public infrastructure projects. #SolutionIsHere

Imagine being able to follow the progress of a road project in your community without relying on rumours or waiting for an official announcement. That's the kind of access DocTracker is designed to improve as it goes live. Public infrastructure should be easier to track.

An AI that admits it doesn't know something is more useful than one that confidently gives you the wrong answer. I think we underestimate how much of intelligence is knowing when you're uncertain.

DocTracker has been deployed and is now in use. The platform is already supporting the Ministry of Works and Infrastructure to track documents, monitor workflows and improve visibility across projects. This is what digital transformation looks like: build, deploy, use, improve

One mistake I see founders make is trying to build for everyone. Instead of creating software for all small businesses, solve a specific problem for independent recruiters, optometrists or accountants. The smaller your market, the easier it can be to understand what people need.