In 1968, John B. Calhoun conducted an experiment named Universe 25, which was done on closed colonies of mice. It was targeted at testing the effects of overcrowding and absolute abundance. Although Calhoun's experiment has not gone uncriticized, it has nevertheless established the possibility of decline in a state of absolute abundance. In other words, the question is not whether or not we should have all that we want, it is what happens if we ever get all we want.
The idea of absolute abundance is one of the foreseeable ambitions and concerns in the development of AI. It is subtly admitted by the doomsday criers and the less perturbed side in the ongoing AI debates. One unspoken divide, however, is that AI—and its impending AGI—are Sino-American forte where all but African countries are hurdled in. While this may in part be because AI companies are not African, did not grow out of, to use the tech speak, African ecosystem, it is more due to the uncritical reception of the entry of AI into African spaces.
And now that major voices are talking about ethical concerns about AI more loudly, what I ask is this: who is speaking for Africa? Either recently or only days ago, the Chief Executive Officers of Anthropic, Dario Amodei, OpenAI, Sam Altman, xAI, Elon Musk, Microsoft AI, Mustapha Suleyman, all admitted one thing that had been said by many critics: Artificial Intelligence, if continually developed without ethical considerations, will result in harm for humans. Since they spoke, King Charles and Bill Gates have also joined them in raising concerns.
You may well ask why it matters to ask the African question now. And to that, I would like to say that the incursion of AI into nearly all areas of human lives, African lives, is a reason. But of greater concern is the unchecked influence that AI has, knowing the vacuity of systems that bestride policymaking in Africa. For a start, no African country currently has a fully enacted, comprehensive AI policy, while a large majority only relies on transitive regulatory powers to oversee AI in their territory—similar concerns have been expressed in The Economist.
The sluggishness is reminiscent of pre-Ebola and pre-COVID-19 days. The problem here is that while the pathogens can be tamed, the intelligence of AI makes it susceptible to a certain kind of escapism. In recent weeks, AI companies have reported rogue behaviors in their models, ranging from absolute deviation to stealthy communication and hacks. What happens if, if not when, this becomes a full-on existential threat? Will Africa rely on external regulations to safeguard its citizens (many of whom are already co-opted in the opium of abundance that AI presents)?
Perhaps to get a clearer perspective of what AI in Africa means, it is best to survey where the world's tech darling is making its impacts. Let us begin with African cultures. The obvious sight is ubiquitous on social media as an algorithm that acts as a gatekeeper of what is seen and heard. What regulatory power exists to verify, if not to direct, the workings of AI algorithmic governance in Africa? That is a question for the political elite more than it is for the ordinary citizens. The infestation of attention rot has torn values apart and installed consumerist capitalist ideals: content creation out of deeply disturbing situations, isolationist preachments, to name but two.
Yes, these are not problems with AI in Africa only, but compared to other places where serious conversations are shaping up, Africa appears to be too laid back. If anything, there are only reactionary responses to the trends of AI, not proactive measures that foresee distant eventuality. The education systems where conversations ought to be led have been silently conquered by AI. The practice of distant reading is more popular than the term amongst students, all with the risks of hallucinated references and non-existent papers.
For now, African leaders, bar Cyril Ramaphosa, mostly see no danger—not likely now, when Donald Trump, who had been unenthused by AI until it was demonstrated to him, has become its evangelist. But that is a dangerous place for them to be. On the one hand, there is the internal danger of desperation for power that AI's deepfakes can sow. On the other hand, there is the external, more direct interference that AI can platform into national systems. What happens when a national electoral database or voting system is hacked with an agent?
The damage that an unregulated AI scientific fantasy trip portends is undoubtedly untold, if anything that CEOs of AI companies say is to be gone by. The damage to the individual and what they represent appears to only just start. Earlier Nobel Laureate Geoffrey Hinton, whose work was fundamental to AI development, had asked: If they're more intelligent than us, will we have a way of making sure they don't want to take over? And we just don't know – we don't know whether that's possible. It is here that more conversations have been concentrated. Sadly, AI use may have already hurt many people with serious intellectual atrophy, a recipe for ease take over.
In Universe 25, Calhoun found that a state of utopia leads to mortal growth, but ultimately leads to extreme mortality when more occupy a space made for less. Everything that seemed to work with four pairs of mice having it all failed when hundreds of them occupied the same space. This goes to show that the abdication of material danger is not a complete removal of danger. The democracy of keeping resources and knowledge within the reach of everyone is a lofty aspiration that undoes the beauty of ambition itself. Humans are made by the processes, complex as they can be, that lead to their cherished ambitions; AI removes that, and that is where regulation must start, but Africa is missing in the conversations.
In the conversations, African leaders should be playing key roles, regardless of where the AI companies are concentrated. AI has travelled the world and is in Africa. Soon, AI data centres might proliferate in Africa's city centres, from Lagos to Johannesburg, Accra to Rabat. How will Africa receive the tech giants? Or, should we again chalk all of this up to postcolonial discourse?
Leave a Comment