Wizards of nothing
I recently watched a TikTok video where two guys were critiquing an episode of the podcast “Diary of a CEO” in which the host Steve Bartlett interviews mindfulness coach Jay Shetty. For the first few seconds of the TikTok video, you can’t tell where the argument is going, as one of the guys formulates what he wants to say in real time. Then he comes up with an unforgettable phrase to describe his view on Bartlett and Shetty: they are “wizards of nothing”, casting “nothing” spells on each other for the hour-long episode. I haven’t watched said episode so I can’t make the judgement, but I can relate to the meaning of that phrase, especially when I go on to LinkedIn. People there have become so good at writing prose-perfect think pieces and coming up with witticisms meant to capture this space between what we know and what we don’t know, but actually say very little. If I didn’t know any better, I’d say they were beneficiaries of a little wizardry. I do know better: research indicates AI-generated content accounts for over half of all the content there. It’s easy to spot and easier to laugh at. But what if being a wizard of nothing wasn’t just the province of podcasters or social media posts, but the guiding framework for a world speeding towards an artificially intelligent future? I’ve spent the better part of a year at Stanford University as a John S. Knight Journalism Fellow thinking about how to strengthen the journalism ecosystem in Kenya and observing the lore of AI build. Being in Silicon Valley in 2026 hasn’t done much to convince me that wizardry isn’t a big part of the story of artificial intelligence. What does this have to do with journalism? Here are my thoughts.
New wizards, same old spells
Having spent the past eight months in Silicon Valley, I’ve had occasion to listen to a number of very impassioned debates about the future of artificial intelligence. The first is a mélange of opinions that amount to the “inevitable AI future argument”. Socially, this presents as worry or excitement about living in a world foretold in AI sci-fi movies. Robot servants, driverless cars, planes and trains. Inch-perfect lawns and minimalist designs. Or tattered clothes and grunge jewellery, AK-47s powered with nuclear bullets, cannibalism and savagery in a world of want, where everything is sepia-coloured. In this future it is the technologist that defines the moral baseline upon which the world is built. Decisions are data-driven and there is a surrender of humanity’s intellect to the power of the machine. The spell being cast begins with the words “AI is coming to replace you”, so the technologists building it should be in charge of directing your future. On cue in late April 2026, the leaders at Palantir publish a 22-point manifesto telling us that the power of the tech company is soon going to be unassailable and state-like, and that they will set the tone for a future after the nuclear age.
Another is a utopia of human conscience. Nature is aided in healing by AI-powered decision-making. Cancer, HIV and other diseases the world has fought against have been eliminated. A universal basic income has been implemented, meaning the entire world has time to pursue pottery and other side-quests. There are no resource conflicts. Helium-3 now runs the world, and a council of states, usually those currently at the top of the totem pole, runs its affairs. The spell here is: AI is coming, let’s keep humans in the loop.
What do these spells actually mean? Is AI coming to replace you? If one looks at what has happened in the world’s largest firms this year alone, it would be an easy answer to give. Tens of thousands have been replaced, their skills automated and performed by one AI agent or another. An entire industry training people for the AI age has sprung up, and the technologies driving this change are updating so quickly it is hard to keep track. Is the value and experience of human beings so joined at the hip to what they produce in the workplace that years, decades of experience, or worse still entire futures, become part of history’s trash heap? How about the other spell: AI is coming, let’s keep humans in the loop? This implies more agency in AI decision-making, so that human beings participate in and take responsibility for the tools they have created. No nominal checkpoints. Real agency that puts us in the driver’s seat when critical decisions are to be made. That hasn’t always prevented humans from making terrible decisions with powerful technology. History and current affairs are replete with these examples.
Both framings are inefficient in describing what the future is going to be like, even if there are signals in both scenarios that will likely play out. The world can get exponentially progressive and dangerous at the same time. Technology can cause great positive change and great harm. Each spell being cast isn’t one that humanity hasn’t encountered before. AI may be more potent, but the questions it poses to humanity aren’t new. What is our responsibility when faced with a tool this powerful? The bet being made in both scenarios is that the historic deployment of capital to build the harness for an AI economy principally by the US and China will yield significant dividends for the winner of that race. Over $300 billion was spent globally in 2025 building out this infrastructure. Hyperscalers like Alphabet, OpenAI and Microsoft are leading this charge, hurtling towards trillion dollar territory in spending towards this goal. Venture Capital, formal banking and governments are all tied up in this, meaning that this bet has to pay off. Failure is too scary to contemplate. So every future AI prospect is framed as epochal, when the actual progress towards milestones like artificial general intelligence aren’t as close as they are being characterized. This all sounds like marketing to me.
Let us not forget that despite its power, it is still a tool currently being deployed under human instruction. If AI is powerful enough to eventually help us re-imagine the way we look at illness, work and the global economy, shouldn’t it also be powerful enough to help humanity re-imagine the AI race itself as something else? Why is it about winning? Government contracts, higher profit margins, lower staff overheads. Even the best LinkedIn think piece is a competition now. Why isn’t it about being, about building systems around the people they touch rather than using those people as raw material? Intent informs how technology is used.
The intentions behind AI
By the time you read this article, a collaborative investigation between my organisation Africa Uncensored, Lighthouse Reports and The Guardian will have been published. It shows how the overhaul of Kenya’s public healthcare insurance system is overcharging the country’s poorest citizens and undercharging some of its wealthiest. An AI-powered algorithm using insufficient training data and a controversial means-testing model incorrectly classified poor Kenyans as living above the poverty line, forcing them to pay higher premiums than they ought to. This happened despite warnings that the model needed to be overhauled. Opportunities to fix things emerged early on but were ignored, because powerful people bulldozed the implementation, likely on the back of promises by Kenya’s president that the healthcare overhaul could be completed in months, not years. The stated intention was to provide better public healthcare to Kenyans. The likelier one was to make a fantastical political promise come true by throwing AI at it, with a goal to win. The result? Kenyans are now having to make choices between paying their premiums and feeding their children. This is what can happen when artificial intelligence is used to win rather than understand who it is helping.
What we should be focused on is whether going forward we can use a tool in a framework that doesn’t run so far ahead of us that it renders our personal or communal needs irrelevant. We shouldn’t build just to be kept in the loop, or to participate in a race we had no part in starting and likely won’t benefit from if won. Why should we in Kenya or Africa deploy AI under the spell that it is coming to replace us at our jobs? What of the already unemployed millions? The millions more coming behind them, who will not find work once they are of working age?
AI and journalism
I’m applying this theory to journalism. What can journalism be if those closest to what’s happening aren’t just interviewees but real participants in the stories told about them and their communities? What if the information they have and volunteer isn’t useful only for individual stories, but can become part of a body of knowledge that helps others make sense of the world? Can it help people synthesize better, and tell their own stories better? What would the tools we build look like? How should they be deployed? What parts of these systems can be monetized for the benefit of the business while protecting the interests of the public? Can the media company and the people who use our services earn more equitably for the work we each put in? All these questions, I believe, point to my focus going forward, and it has more to do with my past than I thought it would.
When I made my first post about what I would be doing while at Stanford, I asked whether, after ten years of running an investigative journalism company, my next challenge would be to build it out as a “house” or a “stage”. The metaphor was meant to capture what I felt journalism’s next challenge would be. Would journalism’s new pursuit be to find ways of protecting itself and its audiences from waves of authoritarianism, traditional or algorithmic, or to build new ways for journalism to flourish? The bet I am making now, informed by what I’ve seen of the AI race, and the thoughtful contributions of many fellows in my cohort is that the future of my organisation rests on two principles: an even deeper trust of the people who have been part of our community, and an approach to storytelling or feature building that is highly collaborative. One of the ideas I’ve been working on is a tool that bridges the gap between static data and actionable insights, by placing human beings at the centre of this framework. I’m calling it “Ground” for now. The problem with most information about Africa is not that it is not enough. It’s that a good amount of it wasn’t built for the people living the realities it describes. Ground is an attempt to fix that — by putting verified human presence at the centre of how intelligence is produced. The closer you get to people, the closer you get to the truth. That’s the principle. The rest is engineering.
My time here took me on an unexpected journey, back to the foundations of my own beliefs. I spent time reading about philosophies that have been the bedrock of African societies, and there I believe I found part of the answer. My ancestors built societies on a relational ontology, the philosophical idea that things exist through their relationships rather than independently of them. Many versions of this philosophy exist on the continent, but the most storied one is Ubuntu. Ubuntu recognises that my existence as an individual happens because of the existence of others in my society. I am because you are. This relationship extends to our creator, to our ancestors and to the unborn, and to the living environment around us. It asks us to act in ways that sustain these relationships, not supplant them for personal gain. A new acquaintance explained it to me this way: the most complex systems in nature evolve from a mutualistic habitat. They are most successful not because one species is trying to win, but because through its being, its making and unmaking, it allows others to exist and to be. Part of me wants success. But I’ve learned enough to know that what I’m building only works if the people it serves feel it belongs to them as much as it does to us.
I’m not going to bet against nature.
